For us IM is faster and also “more insulin”: you get a stronger effect than subcutaneous. So, theoretically you should use less if you were sure it would work as an IM.
But—I a recent IM thread here on FUD, most participants also wrote that the IM effect was not totally reliable: sometimes it works, sometimes not. For us, we use 12.7 mm needles for IM, and out success rate is about 50%. We use thigh and calf, and run for 5 minutes after injection.
@MM2, do you mean less insulin tail? It is very possible, but we never thought of measuring 
Note: the IM thread I mentioned above is here: