Random thought: have you guys thought about algorithmically determining attributes of hands beyond their simple hot/cold equity?
It seems like you could get an approximate measure of hand playability by calculating the distribution of equity against a range on different types of board run outs. Hands that have good playability would be those that have high equity against the range in spots where the player is also likely to think that they have good equity (i.e. if you make a pair you have good equity, if you don't make a pair you have poor equity)
Hands that flop well would be those that tend to maintain equity against ranges on a wide range of boards. The prototypical example of this would be hands like JTs.
Hands like small pairs don't flop well, but have good playability by these definitions.