I think we all agree that top 50 wins matter, but any metric needs to be understood for its inherent weaknesses.
Anyone into "bracketology" knew that the Big 12 and the Big East were going to be all over the bracket back in late December. They are small conferences that IIRC may have both had 7 (or possibly 8 for the Big 8) of teams in the top 50 RPI or really close to it (top 60). So it doesn't matter what happens in conference, the RPI will work out so that more than half its members will have top 50 RPI and hence the creation of top 50 wins.
At the end of the day the a Big East tounrey teams played 56% of its games against top 50 teams. The Big 12 was 67%. The ACC teams in the field may have only played 6 games against a fellow tourney team -- 33%. That creates a huge advantage for top 50 wins.
And this gets back to using metrics properly. Using metrics to seed teams is the fair way to do it -- but you also need to adjust for its limiations.
And once could argue that they did discount the Big East a bit.
The Big 12 had 4 of its members in the top 3 lines, and 5 in the top 5 lines.
The Big East only had 1 in the top 3 lines, and 2 in the top 5 lines.
This is where judgment is necessary. I think they needed to discount even more for the gaudy top 50 records built by playing several mediocore at large teams on your home court. In my case I went 6-8... the committee seems to have went 4-6.