I've never denied that the current click functions are a kluge. It's just the only way I could think of to use the information I had at hand to determine screen locations. With fixes to the initial code it's worked fairly well, and it does have the benefit of not requiring new offsets/etc. if they add a screen location. All we have to do is add that location to locations.txt.
That being said I'd much rather invoke everything directly, if only to solve the persistent problems with custom UI's... and I'm sticking to my guns about any solution to this problem not requiring a specific UI or the default UI.
One big question: is this direct invocation going to work with scroll bars? Current location parsing and clicking does not allow manipulation of these elements, and changing that would be a major fix.