You don't have much of a point clutch, IMO since maps for Crysis all use prefabed buildings proffesional quality maps will be much easier to make, as well as since buildings are fairly simple to model and the such, if INF introduced an additional 50 or so buildings in addition to the number already in the game, there's a huge variety to use. Not to mention that since the buildings are all done, mappers can save weeks if not months building detailed structures and furnishing them, leaving the mapper to focus on terrain, foliage, ambient sounds, and what map specific scenario coding there is. Clutch i don't ever recall you mapping for INF, if you ever did properly do any serious maps you'd know just how time consuming making buildings and furnishing them with doors, lighting, actors, bot pathing, destructable windows, etc. When i made maps or "butchered" them as i did with ACB and Peacemaker, 80-90% of my time was spent just doing the buildings and their interiors.
While the Devs might still be familiar with the coding for UT3, they'll still be severely hampered with what they can do as UT3 is still essentially just an improved UT with better graphics and currently is almost too it's limit, where as Crysis uses a new engine that'll still be able to be tweaked and enhanced for years to come. While Crysis can support adding attachments to weapons ingame on the fly and the such, UT3 will still require seperate models for all the weapon and attachment combinations that INF2.9 is still restricted with. Essentially while coding the game for UT3 will be quicker, it'll still take forever to get new weapons ingame, it'll be just as difficult (near impossible) to add new attachments to existing weapons, and the Devs will still be heavily restricted with what new features they'll be able to implement. Ballistics and penetration still won't be much different than it is in INF2.9. While futuristic vehicles may be provisioned for in UT3, creating modern ground vehicles with realistic handling will be everemely difficult if not impossible. Crysis already has principle of realistic weapons and vehicles in game, and won't require much in the way of tweaking to get how we want, not to mention it already has a destructable world, where as the only destructable things in UT3 will be what the mapper puts huge amounts of time in to figuring out and get working properly, and the amount of destruction will be limited to pre-defined areas. Crysis also has much better AI, and with Crysis (and the current possibility of INF3) being geared towards both a Singleplayer game with a campaign and skirmish maps, and also as a multiplayer game, smart AI will make singleplayer only players enjoy the game more and giving it a much greater replay value, as well as making nights on quiet or unpopulated servers more fun and enjoyable while waiting for players.
UT3 might sill have a loyal community from UT99 and the other UT games, but it's still essentially a tired out old engine and for the most part is supported by older veteran players, while the latter might be so bad, a new game/mod needs newer, younger, 20something year old players for the fresh blood and the new talent that has the skills to produce top quality that takes full advantage of the engine and game capabilities. Crysis is new, and while probably still sufferes from some bugs, it's got a rapidly growing community of all types of people and players, and is looking forward to good developer support in the future, as well as since it's EA thats publishing it (confirm?), you can look forward to at least one expansion pack with more wonderful new goodies to inforporate in the future.
Theres all the arguments i've outlined above going for Crysis, as well as what Nuke has been going on about, and all i see you bringing to the table is a scared little boy afraid of change and whats new, and the one little argument that going with UT3 will mean some familarity for the Devs with the coding.