You edited your post after I started writing (edit: and I just saw you edited this one too, after I posted). But I don't see how it changes my argument. I wasn't disagreeing that lessons can't be learned. But I disagree that the solution is to start from an existing game. That actually precludes it from developing it's own balance and becoming it's own game (and that's not a contradiction with it being a successor to one of the others - after all, FA plays nothing like TA yet is still considered a successor). Your T2 argument, for example. It assumes that because it's not perfect now, it can't be, with the implication that it should be changed to how FA (edit: or any other game) does it. But there's no proof of that. It's not an absolute that the current values can't work with other adjustments. Nor does it mean uber are committed to the current values. They are testing the waters, and that's ok. You can't really be too specific when taking parts from other games, because they play totally differently, and the unit interactions are different. And it makes sense that things work in finished games but not yet in PA, that's why it's still under development.