While clean-up is a bit more onerous, I would recommend only using oil-based clear coatings. I avoid water-based coatings entirely for a variety of reasons, including, but not limited to: 1. their tendency to raise the wood's grain, 2. their tendency to be thicker, thereby obliterating fine detail, and 3. their unverified archival qualities.
Which finish to use is a matter of what effect you want to create. Gloss finishes are to be avoided in modeling because they are not realistic and detract from the compelling impression of reality that is the objective of modeling. "Eggshell" and matte finishes are preferred, but with clear finishes, prepared (i.e. "canned") matte finishes are problematic. Most are simply clear finishes that have been "dulled" by the addition of very fine "dust" material that produces a microscopically rough surface on the dried or cured coating. The primary problem with "satin varnishes" and the like is that it is extremely important to keep the flattening additive in an even suspension in the coating when applying it. This requires frequent thorough stirring (always stirred, never shaken!) and that adds the risk of introducing air bubbles into the coating which will ruin the finish. If the flattening medium is allowed to settle in the coating and the varnish is applied from the top of the container, a gloss finish will be obtained because the "satin" medium is still in the bottom of the can! (Don't ask me how I know this. 
)
If one wishes to use a varnish, the best approach is to use a quality gloss marine spar varnish (also properly stirred, never shaken) with a good UV inhibiter. This UV inhibiter prevents degradation of the varnish due to UV radiation. (Of course, however, for this same reason, no ship model should ever by displayed where it may be exposed to direct sunlight.) Varnish will dry to a gloss finish and when it is fully cured (several days,) it can be hand rubbed with pumice and rottenstone, very fine steel wool, or a Scotch-brite pad, to "knock the shine off" to the level of reflectiveness or lack thereof one desires. Hand rubbing also improves the finish by removing small dust specks, etc., which may have stuck to the tacky finish before it dried fully. It produces a perfectly fair and evenly reflective finish that cannot be equaled by anything that comes in a can. Shellac is treated in the same way. Shellac (as it comes from the can, being thinned with alcohol) will readily soak into bare wood and show no gloss, but when subsequent coats are applied, they will build to a gloss finish. Shellac can be hand-rubbed in the same way as varnish to flatten the finish.
Shellac can also be used to create a French polish finish. This involves hand-rubbing a mixture of thinned shellac and linseed oil into the wood using a cloth polishing pad. French polishing produces one of the most beautiful gloss furniture finishes possible, having great depth and light reflectiveness, but it requires a bit of skill, experience, and a lot of "elbow grease. I don't consider it suitable for ship models because of their shape and surface details. I leave French polishing for flat tabletops.
A very good non-gloss clear coating material for ship models is what is popularly called "wipe-on poly." This is simply thinned clear polyurethane coating. It's applied by simply brushing some onto the surface and spreading it by wiping with a clean cloth or applying to a cloth and wiping it on with the cloth. It soaks into the surface and with repeated applications will build on the surface. (Follow the instructions on the can, as everybody should and few do.) Minwax and Watco make it in the U.S. It comes in clear and in a range of wood stain colors. It also comes in oil (solvent thinned) based and water-based versions. I prefer the oil-based version over water based as previously mentioned with respect to varnish. You can buy pre-thinned "Wipe-On Poly" or you can buy regular polyurethane varnish and thin it yourself with some mineral spirits. The price of the two is roughly identical, so if you buy the pre-thinned "Wipe-On," you're paying a lot of money for inexpensive thinner. If you thin your own, you're getting a full can of polyurethane varnish, not half or two thirds a can of real varnish and the rest cheap thinner. It doesn't matter one way or the other, but if you're as cheap as I am, you'll go for the un-thinned and thin it yourself. I would say that for a "bright" (clear) wood finish on a ship model, wiping on thinned polyurethane clear coating is the easiest, most foolproof method and really looks as good as any other option.
While some favor finishing oils to finish models, I do not. For one thing, all finishing oils are pretty much equal. Real "tung oil" from tung nuts is generally somewhat thicker than linseed oil (which is the same thing as flax seed oil in the health food store,) but not much different otherwise. The only thing "teak" about "teak oil" is the name on the can. It is not made from teak trees. They just call it that on the can and charge more for it. Oil takes a long time to polymerize ("dry") and is messy stuff to work with, particularly on a detailed surface like a ship model. ("Boiled" linseed oil isn't actually boiled at all. It has Japan drier added to acceleration polymerization so it dries somewhat faster than "raw" linseed oil, which can take weeks to "dry.") Other than that, wiping on thinned finishing oils will yield a finish comparable to "wipe-on" thinned polyurethane coating, but likely take much longer to "dry."
Two points you may also wish to consider are:
One, any glue residue on the surface of the wood (E.g., thinned CA or PVA adhesives soaked into the wood) will cause an uneven coloration of the glue-coated wood when a clear finish is applied. A stained finish will similarly be uneven in such instances. Unless the wood surface to be finished showing the wood's coloration is completely free of any glue, any attempt to finish it bright will likely result in disappointing, if not unacceptable, results. One way to test the wood surface for this impediment before applying a finish is to wet the bare wood with a fine spray of water. The wet surface of the bare wood will show any unevenness in coloration in the same manner than it will show when a wet coating is applied. While this may raise the wood grain and require a fine sanding before final finishing, it's worth the effort and cheap insurance against a finishing disaster. If the offending glue cannot be sanded out of the wood without ruining the fairness of the surface, there's nothing for it but to abandon hope for a bright finish and paint the wood instead.
Two, although we have no historical evidence of what the two Fifteenth Century vessels you are modeling actually looked like, it is safe to assume they were never finished "bright" with varnish. They were undoubtedly "tarred" with what today is called "boat soup," a mixture of pine tar, linseed oil, and turpentine. While the initial coats of "tar" would have been thin enough for some wood figuring to show through, in a very short time, additional coats and the effect of weathering would have yielded a very dark brown color that, from a scale viewing distance, if not up close, would have appeared to be flat black. (And to this I can attest from my first-hand experience with one of the full-scale "replicas of Pinta.) Colored paint pigments other than whites and brown shades, were very expensive in the Fifteenth Century and only used, often at the captain's own expense, for decorative accents and then only rarely on working merchant vessels. (Blue pigments, then made from crushed lapis lazuli, were so expensive that they were only used in fine art works.) A "bright" (clear) finish on ship models is a stylistic choice. In most all instances, it bears no resemblance to the actual appearance of a vessel, but some modelers like it. That said, there's nothing wrong with a model that's painted instead of finished bright.
I agree with most everything Bob Cleek said, but I would like to offer another perspective.
We are finishing our models more for aesthetics than for protection (assuming your model will not be displayed in a bright sunny window or outdoors.) Any solvent based product (oil, thinned varnish, shellac, etc.) will soak in on the first coat and give you a relatively flat finish. If that's what you desire, you're done. If you want slightly more sheen, put on a second coat. I would go no farther than that for a number of reasons.
The more coats you apply, the more it will become necessary to scuff the surface to allow subsequent coats to adhere. That can be difficult if not impossible on some areas of a delicate ship model. Plus, if you don't get into all the nooks and crannies, they will stick out like a sore thumb.
A rule in the woodworking and furniture industry (in which I work) is that ALL wood surfaces, inside and out, be finished equally, lest uneven moisture incursion cause uneven wood movement. Moisture
will migrate in and out of the wood with seasonal changes. If you don't think small pieces of wood can move much with humidity, just G00gle "weather stick". If you've built your model in the typical fashion, you have planked and decked it before any finish is applied. That means there is no finish inside, moisture will permeate faster on the inside than on the outside and cause uneven wood movement which could lead to warping and/or popped glue joints. The fewer coats of finish you put on only one side, the less this is a problem.
If you do use metal wool, use brass or bronze wool. Small particles of the wool will inevitably break off and lodge themselves in the wood grain. Steel wool will eventually rust and show up on your model, the others won't.
Finally, to see any glue spots prior to finishing, I recommend a solvent rather than water. As Bob mentioned, water will raise the grain requiring sanding. Again, this can be difficult on a model without losing detail. Use "real" mineral spirits, alcohol or acetone. I usually use alcohol because it stays long enough to see the spots, but still evaporates quickly. None of the solvents will raise the grain like water does.
