Here, I wanted to look more close at Monk's example of KQJ9 on a board of QT5. As opposing hand, I am taking the best pure draw I could find, namely AKJ9 (which has to play regardless of what happens imo).

Here's what I came up with:

First, if both the board and the draws are rainbow, then KQJ9 is a big favorite at 1.9:1.

The odds against AKJ9 are also exactly the same if we change KQJ9 to KJT9, pairing the T rather than the Q. However, with regard to the original question I was talking about, I think it's much better to have the Q pair than the T pair because it makes top set less likely, and top set also has to call (and is favorite). If they don't have top set, then you're against putting a lot of pressure on middle set and 2 pair. The latter really has to fold imo, and probably against the range I'm trying to work out for raises, folding middle set is a pretty good idea as well.

But things change significantly on twotone boards. There, if AKJ9 has the nut flush draw, too, there's no way I can find for KQJ9 not to be underdog--although not really dominated underdog, I'd say. Anyhow, if AKJ9 has nut flush draw and KQJ9 is rainbow, we're looking here at 1.4:1 for AKJ9.

If KQJ9 has K-high flush draw but AKJ9 still has nut flush draw, then it goes down to 1.3:1, but AKJ9 is still favorite.

My conclusion: On rainbow boards KQJ9 is a clear push hand on that board.

But I think on twotone boards, you need not only TP but also a big flush draw to push. If, for example, you have the K-high flush draw, you're in better shape vs. AKJ9 AND you have a lot more draws in there that your hand dominates. With no flush draw at all, I think KQJ9 unfortunately becomes a fold (?) on that board--at least to a pot-sized bet.