In the abstract, fixing the root problem >>> maintaining a shitty internal wrapper that has to get reworked every time upstream gets fixed. Not fixing the wrapper repeated is absolutely business value, because not having a half assed fix of an implementation means fewer bugs in terraform plan and apply, which means less time messing about with CI because the stupid thing broke again.

This is of course in the abstract. I know neither the quality of the wrapper you wrote, nor how long it would take you to do clone a repo and write some code for any upstream fix, given all of your competing priorities.

My underlying point is that not fixing things properly has a cost all its own, and wasting time with a half assed solution can cost more than is immediately obvious.

It's impossible to say in the abstract if it is more efficient to actually fix the root problem and be of more business value vs shitting out some wrapper script, it depends on the downstream effects of said wrapper. But I've definitely avoided doing an upstream fix and wasted countless company resources getting a wrapper working when I could have rolled up my sleeves, done said upstream fix a year or two earlier and overall saved the company money.