My coding efforts were during my PhD and postdoc days, using Fortran and an arcane version of it called Mortran. The largest effort was writing a 6000 LOC program that seeded and guided "elementary" particle interaction Monte Carlo simulations.

Since those efforts only lasted until I finished that postdoc, or about 8-9 years, and I never had to code as part of a team or deal with agile or OO programming, I never burned out from those efforts (though I certainly had far more than my fair share of supervisor-from-he'll experiences...).

However, two of my three kids have been software engineers for over a decade, and I see evidence of some of the problems you mention. Overall, things look super shiny from a distance, but the details can often make coders' lives miserable.