As just one more conception of a Star Trek MMORPG, I think there's room somewhere in between "an entire game ... needs to exist before player ship interiors become useful" and "nothing more than a ship-based recreation of a Trek episode".
I favor designing player ships with a few key interiors. Absolutely there is a cost to doing so; the question is whether the potential benefits sufficiently exceed the likely costs to make them worth accepting.
A big chunk of my "yes" answer comes from the following beliefs I hold:
1. Serving aboard a powerful and complex starship is one of the two or three most crucial aspects of all of Star Trek. Hub ships with interiors and operable systems will help offer that in a Star Trek MMORPG, but not enough if most players spend most of their time aboard player ships. To maximize their value, key license features need to be implemented where the players are... and players in Star Trek Online will, I think, be predominantly in player ships.
2. Gameplay consisting of working with a few other players aboard a vessel with segmented systems would positively distinguish ST:O from SWG and EVE, its main competitors in the science fiction genre. Those games are mostly about single-player fighter craft, giving Star Trek Online a legitimate chance to be the only mass market AAA online RPG offering a richly detailed multiplayer shipboard experience.
3. Interacting with the other players on board a ship solely through an outside-the-ship "viewscreen" metaphor will not deliver the same You-Are-There feeling provided by a 3-D interior model that allows characters to interact with other characters as avatars. Tactical efficiency is not the only thing that matters in designing the character interface model for a game.
Overall, I think a measured selection of features that are likely to pay off their development cost is exactly what ST:O needs, for the question of player ship interiors as for every other design question.
Others have expressed their opinions on what would be the best measure, and that's fine. For me, that measure comes with implementing a few key interior locations -- say, Bridge, Engineering, and Sickbay -- aboard every player ship big enough to reasonably have such locations. Let those interiors house controllable ship systems, and populate each ship with enough NPCs of the right type to offer interesting content for each location based on the ship's current mission.
At that point I think you have a game design that makes effective use of the license.
