The 'problematic benefits' for micromanaging gourmand, those discussed in this thread, seem extremely marginal to me
This doesn't mean there isn't a real problem. You can absolutely choose to not address the problem (I mean, you're the ones in charge of the code)--and I acknowledge that it's a small problem, and perhaps you would better spend your time on other things--but it's still there, so please don't try to pretend that it's not. The benefits are absolutely problematic, because they both fit the definition of grindy behaviour that crawl tries to avoid, and additionally gourmand has a charge time that ostensibly tries to avoid this specific problem although as I've said it doesn't actually fix anything.
Saying "I don't think it's worth my time to fix and I think gourmand existing has benefits" is okay, but I'd prefer if you actually just said that clearly.
If they needed to be weakened further, we could add a random factor to chunk rot times
Unless you make the minimum time 1 turn, this does nothing except move the window. I don't know how long chunks take to rot now; let's say it's 100 turns. If you make it 20-200 turns instead, then instead of waiting 99 turns before I eat (ok realistically more like 95 or something), I just wait 19 (15). Exact same behaviour. You can still randomize chunk rotting timers for other reasons, but it does nothing for gourmand.
---
Anyway I would support taking gourmand off setting you to (near) starving, since that does fix the problem, is pretty easy to do, and apparently already has a patch.