To give this a little more thought, I think it’s helpful to look at the correlations between the categories. So this correlation matrix is based on the projections that I’m using for all players with a value greater than $0:
Let’s first compare FTM vs FT%. As you can see, the correlations between FT% and the other stats is very high or moderately high, with the exception of STL (r=0.03). On the other hand, FTM is very weakly correlated with BLK (0.05), FG% (0.03), and 3PT% (0.06).
In categories, you want some statistics that don’t correlate very well; otherwise, you’re essentially just measuring the same thing over and over again. For example, 5x5 features HR/RBI/R (which are highly correlated with each other) but also BA (moderately correlated to HR/RBI/R) and SB (weakly correlated with everything). So simply put, FTM is a more interesting category than FT%, IMHO.
In terms of 3PT% versus 3PTM, it’s a similar story: 3PT% is weakly correlated with STL (0.08) and FTM (0.06) again. Meanwhile, 3PTM is correlated with pretty much everything.
So basically, if you had FT% and 3PTM instead of FTM and 3PT%, you’d have fewer tradeoffs to consider when constructing your team. And that’s really the whole point of playing categories rather than points, IMHO.