If one does not care about lisp, then one would likely not care about lisp machines.
And there is memory left in the community--just not everywhere. I don't recall Ted Nelson's stuff saying that we are doing anything right, but I could be mistaken.
There is a bit of cognitive dissonance as well--if you like your current IDE and you are programming in java and it solves refactoring nicely for you, then you might not be likely to look elsewhere for good IDE ideas.
And with non-lisp languages, it is a lot harder to get the IDE to "understand" the program in the way that lisp IDEs do, even slime/emacs combined.
I would think that a side-by-side comparison of "modern" IDE and lisp machines would favor the lisp machine.
That being said, I recall that pg once wrote that they spent too much time on those fancy features.