I can see it both ways. Paying 5 bucks for an add-on that's reputable and not having to think about it again may be preferable to shipping your own. It's not evident to me that one-shot-esque LLM programs will ever work as expected, since they're limited by the amount you have to specify, and then maintain as new issues arise. It's the type of work that a lot of people will be unwilling to put in.

That said, I also think that public reviews will start getting recalibrated by the users themselves. People will start noticing that programs are unreliable or lack features which are seemingly simple to implement with an LLM. The quality of an app may very well depend on how hard it is to replicate, otherwise, why not ship your own?