I just had a blinding flash of the obvious as I was slogging through gorp of VoIP in my job: the sorts of things I do generally aren't respected by the Computer Science academic community because they're not simple enough; they're not elegant enough.
It seems that only simple things generally receive high praise by academics. Or perhaps it is that academia is about simple things that follow simple, elegant, knowable laws.
We live and work in a world of taming immense complexity. Much of it is unnecessary, but it is necessary to tame it and make it work properly. But my advisor had no respect for this type of work . . . and I have a hard time convincing myself that it is respectable.