In the web interface, if you enter a name that is isn’t a recognised taxon, the first thing it does is give you an option to “Search external providers”, and if that turns up nothing it then asks if you want to “Use what you entered as a placeholder”.
If you do that when suggesting an ID for an existing observation it only offers the first step. If you want to describe the observation with a “placeholder name” you can only add that in a comment.
Maybe there is a need for that to include a little “what is a placeholder name” help button with a brief explainer (and I won’t rehash what people have already pointed out may be shortcomings in the apps re this) - but “Placeholder names” are a fairly widely used term of art, that unlike some species common names, fairly unambiguously always mean the same thing.
We have placeholder names that we use for discussing species that have not yet been formally described but are well known enough to be commonly discussed. We have placeholder names for proposed new species when research is presented that suggests an existing species should be split but that new name has not yet been ratified.
We have placeholder names in common use “John Doe” for a body we don’t yet know the real name of.
The ncbi.nlm.nih.gov taxonomy faq says:
“Unpublished taxonomic names are not allowed in NCBI records. Consequently, placeholder names are substituted until such time as associated proposed novel names have been published. It is made clear to submitters that it is their responsibility to notify NCBI when a new name associated with their sequence submission(s) has been published.”
The gtdb.ecogenomic.org faq says:
A strain identifier is used as a placeholder for the genus name when there is no existing genus name and no binomially named representative genome.
etc. etc. and many other fields use this term the same way.
So I appreciate the problem of this being jargon that may need explaining to people not familiar with the idea, and all the problems that flow on from that - but this isn’t something that inat staff or programmers have made up from whole cloth with their own special interpretation, it is a recognised term of art, and necessary transient situation anywhere a living but strict taxonomy is used.

(I do feel the placeholder issue is a sperate issue than the title topic of this thread, but I can’t exactly post this reply to the closed feature request. I’m mostly hoping Tony will get more notifications that he has been quoted.)
Yes, I’m nowhere saying “placeholders are perfect, don’t mess with them” - just that they are an entirely separate problem to the question of whether having users add coarse IDs is a Good Thing. If there’s a bad interaction between the two, that’s an implementation detail (which may need improving), not a reason why ‘coarse IDs are bad’.