That circuit could be right in that the payoff is often not actually worth the effort, when viewed objectively and holistically. For example, if you literally spend two months working on end-to-end tests and it in fact does save you three whole weeks of work in terms of debugging or whatever over the course of the next six months, that doesn't add up.
I think there are some common extremes. On one hand, the business side often forces engineers to take on tech debt that really is the wrong tradeoff. On the other hand, many engineers may spend time creating a sort of idealized factorization and large test suite that in the end really isn't going to pay off. And part of that is actually just because they are worried someone else will find a nicer code organization or another test to write and judge them for not doing it.
I would say this is actually the biggest problem with tech debt, the damage to morale that it causes.