The opposite experience was had in semiconductor manufacturing. Not a soul in those engineering offices would hazard the remotest assumption about a problem until many hours of confirmation occurred first. No one wanted to be the reason something got even more complicated.
Compare to received startup wisdom: reduction is fine, often 80:20 is good enough, and failures are expected. You miss all the shots you don't take. Sometimes moving in the right general direction with wrong perception is the best path to progress. Learning is a goal.
The percentage of failed ventures which can be directly attributed to reduction-related problems of founder perception is probably very small.
It's more like "working in enterprise" (aka a large, already somewhat successful business with all the factors that entails) than startup land where problems can be "unpicked" more easily because the system does not have as much state, history and inertia.
Nearly every problem involving people or the environment is going to become a wicked problem. The problem space is just so big and you can't really test at the scale you'll have to implement on.
And it's hard to know what works and what doesn't. Because sometimes things can get better due to something else and your solution just happened to be implemented while that was happening. And it's not like you can isolate either.
Essentially, you're always testing in production.
Plenty of problems are well-defined. Even the ones involving people.
Just because people create situations that are wicked at the macro level, it doesn't mean that most problems reside there. If most problems were wicked, society would never evolve.
Also, just because a problem isn't completely solved, it doesn't mean progress isn't useful.
But figure that we don't have definitive ways to effectively teach people. We have ways that work somewhat for some people, but disseminating knowledge in ways people can understand is a wicked problem. Anything you can think of, we have probably tried. And there is still some segment of the population that it does not work on. And those methods will work on those people in other cases. But not the ones those methods worked on in the first place. So obviously the method is not general, it works in some cases.
People make problems wicked. Because every problem becomes incredibly multi-faceted as you are dealing with so many different fields at once.
And it is the navigation of these wicked problems that is responsible in part for evolving us. A lot of societal progress is the result of attempts to solve wicked problems.