I've thought about this a lot as a composer and as a musician looking to fall back on a more improvisation-based process. The thing that's been on my mind is that we rely on technology ... a lot. As the OP mentioned, lots of notational software gives you constant playback/feedback when you input notes. It's weird that we're the ones supposed to be doing the hearing, and yet, we've got our computers to do the hearing for us in most cases. Shouldn't we definitively know that what we're inputting is the right note, based on a combination of internally hearing functional harmony and knowing what components of theory give us -those- sounds?
(I personally think that I could probably start doing things this way right now if I wanted to, but it would be painfully slow and fraught with self-doubt over the intrinsic 'correctness' of the notes that I think I'm putting in ...)
Don't get me wrong ... it's good to hear what you're doing as a means of affirming that you're taking your work in the right direction. But a lot of the greats of the 20th Century never had someone constantly factchecking their work as they were making it. Maybe -after- completing a manuscript and getting it sent off to the press to be mass-printed would they have editors swoop in and address any little oddities. But so the stories go, guys like Bela Bartok would go out into the country to record folk music on gramophones, then transcribe the recordings to sheet music without needing to consult the piano for every single note.
I guess the pressing analogy is this: a painter shouldn't need to check his palette to remember what red looks like. Musicians/composers shouldn't have to check their instrument (or computer as the case may be), either.
Ron Gorow's got a great book that pretty much addresses these points in full, if anyone's interested in developing their ears. You will get through the first part that gets you to sing intervals with probably not that much difficulty, if you do the drills properly and you're willing to take the time and learn. Then you'll get to the first section on chords ... then cry. I'd know ... I'm still stuck on the first part of the chords chapter. His philosophy is that modern education has too much of an emphasis on major/minor/extended distinctions, and that this just makes the simple problem of dissociating simultaneous groups of notes into a overly complicated textbook classification problem. So basically he has you do things like tetrachords. Not for the purposes of identification, necessarily, but perceiving chord "depth" and intervallic distances within the chords.
TL;DR: music's hard, and not sucking at hearing it is harder.