I guess I’ve been summoned three times so I have appeared. I agree with others that placeholders aren’t particularly relevant to the original question. To answer the original question, I think that adding coarse IDs has significantly more benefits than it does drawbacks. I do it all the time and see observations get refined very often. And even if that ID doesn’t get refined, the observer may have at least learned something, even if it’s that there’s a separate category for vascular plants, or that harvestmen and spiders are two different but related orders.
As for placeholders, I don’t think the way placeholders are treated will be changed. The term “placeholder” does point to something fairly ephemeral. The observer can add something to the obsevation notes/description if they want to make something permanent. The upcoming mobile app will also have offline suggestion options, so people will be able to add an actual identification of some kind to their observation if they’re in the field and lack connection to iNat’s database, and personally I most often see placeholders used in that situation. Hopefully these changes will reduce the need for them, and reduce people using them by mistake. We’ll have to see. But even if the placeholder is removed, the observer added it and should be able to remember what it was if it was that important. I usually see things like “plant” or “bee” (in various languages) and I don’t see much value in those for the observer.
I read thousands and thousands of emails sent to iNaturalist every year, and placeholders are basically never mentioned, so I doubt they way they’re handled bothers most people. If it was, we’d be getting more complaints, because plenty of people send us criticism and suggestions about all manner of things.