Yes, stellar mass may be the better approach. The number of stated masses and the number of star types are about equal to about 500 lyrs., but overall, there are 4352 masses stated for stars, but only 1877 star types.
I'm a little surprised at that given the importance of star types (i.e. temp.).
However, the number of stars with a given radius is about as extensive as the stellar mass. Thus, it shouldn't be that hard to get a reasonable luminosity result, thus eliminating reliance on the limited number of star types. Perhaps I should use star type first then the other two parameters secondarily for HZ determinations.
Is this what you do?
Of course, your result likely enhances my view that Sun-like stars are more prolific in having planets even if some fail to survive.