In general I like the idea, it adds depth to gameplay and introduces new strats. Although I think it needs to not be over-complicated and should be obvious what is happening.
Simple to learn, yet hard to master. Generally whole good game interface - in wide meaning - shall be simple but have potential, let a player become better and better in using it in right ways, right times, and align all the pieces to form his/her own unique style. Maybe I rushed a bit with the “unique style”, because some good practices will repeat, the thing is - good design of game would allow them not to be a panaceum for everything.
Story - I was once playing a RealTimeStrategy game and it was fun until I’ve found out there is a single always-working victory pattern to follow mindlessly. Camp yourself with defences heavily, then make alot of cannons and rush the enemy out of the map. Finish. It worked for me on the highest difficulty level even. Maybe it is due to being used to the game so much nothing needs rethinking - so much the worse. I’d like the tactical dimension of the DB to require some cleverness and never ever allow the same tactical pattern to defeat everything no matter what.
Complication - easy to grow it when adding more and more complexity to the base(the game, not interface - let that mean that for a while), yet there are ways to provide effects that are due to standard-routine choices. Just like class-choice. User has only to chose his class. New player will have no problem with the additional feature, yet as he will get experienced, he will choose his class also regarding the effects of the structure.
As this comes simple, yet gives some new possibilities to shape the team consciously, not only in terms of we got that many medics, this many engs… but some buffs or even changes to the game-ground - the battlefield(like for example more field-ops = better communication(faster information about side objective status, maybe main even. You name it). What could it influence, is a good deal of a question. What do you think?