I used to think this, and I’m sure there are plenty of bad architects who add net-negative value, but having worked on some extremely difficult systems as an IC, I would have given anything to have future me able to hand down a scalable architecture from on high, vetted by past experience and domain familiarity.

Not having that, I developed the knowledge myself through trial and error, but we would have saved a lot of time, money, and stress doing it right the first time.

In general, I think this kind of “architect bad” take underestimates the cost and the stress of being responsible for a system that ultimately isn’t a great fit for the domain, and needing to balance hacking another fix onto it vs migrating to what now know is the right thing.