the 3 point shooting rating trilogy has ALWAYS been spd/per/bh. nothing in the new engine has changed that. of course iq and stamina matter (for fatigue level), and presumably ath plays a small part, although i wouldnt be that surprised if it didnt. its almost certainly a distant fourth, at best. ft% matters too, but it matters least of all for non-scorers and for heavy 3 point scorers, so i usually am not very worried about ft% in my pure 3 point scorers.
the offenses vary in the ways you can build successful 3 point shooters. of course, they are all 95% the same, but there are subtle differences. for example, in flex you can focus on spd/per and omit bh with greater success than you can in motion. in motion you can seemingly build as effective 3 point shooters, even in high d1, with slightly lower per ratings, if your spd/bh are great. ive had guys in the 80s every bit as good as guys in the 90s, as long as their spd/per are up to par.
the way shooting works, it seems to me, in a simplified view, is that you basically have a 2 step process. the first establishes how open you are - which is really where spd/bh come into play (as well, i would assume, as team passing). the second is where you take the shot, and that is primarily per based.
you can see evidence for this based on the way various players perform against different levels of opponents. when you look at teams with weak SOS, you often see these high per, lower spd/bh guards doing quite well. don't be fooled - those players will suffer mightily against strong defenses. when you look at those same players on great teams, they really dont get to shoot much at all, because frankly, they suck. on the other hand, guys with high spd/bh and solid per can score efficiently against quality opponents, albeit in a supporting scorer role, maybe a 10ppg level, not your #1 guy. however, these guys dont get that much better against crappy competition, like your lofty per players will.
the level of play you are at, and the quality of defenses you face, absolutely effects the extent to which different ratings help your guys score. i believe this is why there has always been a substantial amount of disagreement among the community as to the importance of lp and per for "normal" bigs and guards. the reality is you can lean on lp and per heavily, sacrificing other ratings, and get away with it, when you play shittier teams - but try to compete in high d1 like that, and you will get crushed. you simply are going to be decently open anyway, when you play against poor defense - but against the good teams, you have to really create some separation to get any kind of a decent shot off - your per is largely irrelevant if you can't get good looks.
to try to concisely state this so coaches can take something away from this - if you are in a rebuilding mode, not competing at a very high level (even up to a solid 2nd round NT team in d2/d3 may qualify here), then its per by a substantial margin that is most important, followed by spd/bh. but if you are trying to compete at a high level, you are going to need quality spd/bh to get good enough looks to take advantage of your players' per. per is still the single most important rating for 3 point shooting, that is always the case. however, in high end play, if a guy doesnt have strong spd/bh, his high per is largely wasted, and you cannot consider this guy an offensive leader for your team. you can sometimes redeem these kinds of players by bringing them off the bench. the point i guess im making at high level play is, if you want a guy to be a high end scorer on your team, hes got to be strong in all 3 (of course exceptional in 2 allows subpar ratings in one of the others, but you know what i mean).
also keep in mind great spd/bh players with poor/mediocre per can still be valuable scorers - your mid tier type scorer, 6-10ppg, where efficiency, not quantity, is key. meanwhile, great per players with poor/mediocre spd/bh are more or less useless, at high level play. those guys can still be pretty useful in lower end play, especially off the bench.