I think the problem is there isn't a clear delineation between "traps" and "meaningful improvement."
Take Signal for example - early days they had a ton of success with a core group of users, in spite of a number of product warts. Ever since then, they've been making usability improvements to lower friction and appeal to more and more marginal users. Is that good or bad? Based on the hn threads I've seen, it seems like the jury is pretty mixed?
Signal is a special case because it's a generic messenger app - meaning it's target audience is legitimately everyone, and with its philosophy and E2EE in general preventing interoperability, it has to capture everyone. It has to be optimized towards a Marl, or else it'll fail to network effects.
Some software is like that. Most isn't - but the enshittification culture is affecting it all the same.
I don't think signal is an especially special case. While it may be clear and obvious to you which apps are which, I think that the reality is that its never a clear delineation.
Take Signal for example - early days they had a ton of success with a core group of users, in spite of a number of product warts. Ever since then, they've been making usability improvements to lower friction and appeal to more and more marginal users. Is that good or bad? Based on the hn threads I've seen, it seems like the jury is pretty mixed?