I've won 4 NC in 5 seasons in Carpenter, and one additonal NC in Bava. I'm not a techno-geek or a programmer and I don't pretend to understand the game in its entirety. That said, I do think I have some grasp of 3.0. I also think:
1. You don't have to run a counter-intuitive game plan to beat SIMs. I don't pass out of the bone or run out of the gun. And I'm a perfect 12-0 against SIMs between my two teams this season and I understand why.
2. Talent does matter. Not as much as it should, but it does. If I run the same gameplan against BCS, midmajor, and D1AA, D2, and D3 competition, the score in my favor widens at each level.
3. Trying to fix the beta by manipulating final results creates new problems. The first two seasons of the beta, running was king. That was overcorrected, and passing 100 times became the vogue. That was addressed by manipulating QB fatigue, which resulted in another overcorrection.
4. The debate about whether an 80/80/80 player should always be successful against a 79/79/79 player is a red herring. The talent differential there is minimal. the more salient question is by how much should an 80/80/80 player outperform a 75/75/75 player? Or a 70/70/70 player. Permutations and results should be based (I think) on scales of talent disparity - how different is 3 points in cores? How about 5? Or 10? Meaningful differences should define the game and results.
5. Just like 1.0. there are some holes to be exploited in 3.0, opportunities to create a winning strategy that stands up game after game after game. Winning in 3.0 isn't really a coin flip. But 3.0 doesn't give coaches the same ability to choose a winning style that 1.0 did. Good coaches in 1.0 could (and did) build teams that could win running or passing. I don't think that's true in 3.0 yet.
I know there's lots of unhappiness and 3.0 isn't meeting the expectations of many of the beta testers. Setting starters is a pain is the butt. It's hard to understand why in reviewing the debuged PBP some advantages are strong and others nonexistent when the attributes don't match up to those "buckets". 3.0 isn't close to being a finished product, but there are some building blocks here that could bring back much of what I liked about 1.0. The truth is, no game can have 100 developers. Some of the testers - like me - want 1.0 back or something like it. Others want a game more like 2.0. No one game can be both, so some block of coaches is bound to be disappointed.
From a strictly commercial standpoint, it seems 2.0 sucks. Worlds are empty. Yes, reward points tanked but the coaches didn't leave because of that - they started playing without the points and would have continued, I think. The game that built GD and led to full worlds and expansion was 1.0. In my opinion, the more 3.0 miirrors the 1.0 experience, the better the chance for success and to fill the worlds again.