What I like about the tech debt metaphor is that it extends to the concept of “interest” so cleanly – you can literally point to a piece of work you’ve spent some amount of time on and say “about 90% of that was interest on piece X that can be improved by Y, which should take Z time to pay off”.
I’d be curious to hear the discussion of “knowledge debt” taken in this direction, as well – there are, of course, unknown unknowns: at which point does one say “had I known what Python’s meta-classes are, they would have been very useful here”? After you’ve paid off the debt, i.e. not how financial debt works.
However, I can say “darn, if I took that class on Coursera, I could probably have been significantly more systematic about my debugging of this scaling issue”. That seems a little more wishy washy, but perhaps it doesn’t have to be?