My only mild opposition to that would be the fact that you couldn't directly compare Season 1 rankings to Season 2 rankings. Not unless the roughly same number of predictions happens as Season 1, which would be a bit of a challenge given the usual workload plus confidence rankings
But otherwise...