Well the Maximum call stack size exceeded seems like a bug to me, would be interesting to know conditions that reproduce it (isn’t directly tied to Background Task error imo).
And regarding the LOC threshold:
I feel like this needs to be thought through first, too early to jump to solutions right now. ^^ My reasoning: the amount of LOC purchased (and therefore the $$ paid) could be considered a private matter by the org admin, one which the org admin might not necessarily want all project admins to be aware of.
All in all am not saying the current message is ideal (understand the unclarity @gregpakes, and indeed right now you’ll have to iterate gradually), however I would suggest we just track it from there and discuss the appropriate solution ad-hoc. @Fabrice_Bellingard: sounds fair to you ? (@gregpakes: am also thinking that it might be worth for you to make a dedicated [Deprecated] Suggest new features out of this, to keep it distinct from the initially memory-focused discussion in this topic, and track it from there).