I mentioned this already, and forgive me if I sound like I'm repeating myself, but if it's going to be like a year of work no matter which engine you pick--okay, maybe you can save a month here or there or get some slightly better feature with this or that engine, but still it's a massive undertaking whichever one you pick--the number one reason I thought it'd be best to pick TDM wasn't really because it's the best option for multiplayer, but if it did have multiplayer, then you'd have over 100 maps to play coop style right out of the box. And my experience with T2MP is that if a FM is great for single player, it's almost always also great for coop, and I imagine comp too (because players just take the place of already existing guards, so the level is already built with thieves vs. guards in mind).
I can't tell you how much fun it is to coordinate how you're going to rob a mansion together with a friend, and when the unexpected happens, you can run to your friend's aid. When I'm playing single player and I get caught, I'll be tempted to restart, but if there's another person there, you want to game on because either they're coming to help you out, or you're watching your friend get chased across the grounds and it's hilarious. And so many of our maps are so good. I can't imagine someone spending a year making multiplayer on another engine and we only get 3 maps, when we could have all of our great maps to play with a friend! It may be extra work on the multiplayer side, but so worth it IMO.