And trust me, I've had my issues with using some of these forward looking/prediction based models for tourney seeding. I'd rather select and seed based on the best season, not the best team.
But no matter how you feel about that, this isn't a power ranking, or anything like that, it's meant to be a very simplistic, forward projecting model, so yes, KenPom or NET or whatever is a better comparison to the sole purpose of what the tiers are trying to do.