TL;DR: it’s not my intention to force you to do or use something! I just wanted to help, and understanding what exactly you’re trying to accomplish (and in what way exactly are current tags insufficient) is a prerequisite for that. You can always use any tags you like, but unless they’re good ones, others might not. Good luck!
Well, OK, surface is “just a surface”, but it does have some properties (one of which is “looseness”), which is why I gave specific examples of mud and sand (and there are other examples, e.g. fine_gravel is loose one, while compacted gravel is non-loose one, as is surface=asphalt).
How does an a surface=mud which is not “loose” look like? Or an “unloose” (firm? compact?) surface=sand? What is surface=loose_sand and how does it differ from “regular” surface=sand (which looks very “loose” to me)?
Uhhh, I don’t follow. Bicycles are vehicles. vehicle=* tag applies to them.
Maybe you meant bicycle vs. motor_vehicle? But motor_vehicle also applies for e.g. motorcycle and mofa, which share many of bicycle issues. Maybe you meant 4-wheeled motor vehicles only? But even there there is a huge difference between say atv=* and hgv=* – the latter will have even more issues on that sandy beach than mtb=*, while the former might just work fine. Much of that “looseness” is as important (or even much more important!) to other vehicles as it is for MTBs, so it does not have sense to be under mtb: prefix.
Well, I am trying to understand what exactly you are asking about. Do you have picture of that “loose sand on a road being grade 2” so it can be better understood? Picture is worth a thousand words! And what you mean by “grade 2” there? mtb:scale=2? or tracktype=grade2? (And of course those are different!) Or do you mean some sand being blown on the otherwise asphalt road? Or something else entirely?
Is it using OSM data or something else? Because I don’t see “persistently_muddy” as widely used in OSM (there are 0 matches for key persistently_muddy=* and 9 matches for surface=persistently_muddy worldwide. Or did you mean regular surface=mud there? Or something else? Perhaps some non-OSM data?
Uh, I’ve tried, but (not speaking Finish) I can only recognize those icons of fallen trees to probably indicate obstacle=fallen_tree. But no idea about “persistently_muddy”, which is only one related to this thread? Could you be more specific here?
Perhaps I have different understanding of word “persistent”. What you seem to describe would be “temporary” as I understand it. I’d only mark something as surface=mud (i.e. persistent mud) if it is actual “mud” at least 3/4 of the year, preferably more. Otherwise, I wouldn’t map temporary features (and even if I did, it is quite likely other mapper would overwrite that surface=mud with e.g. surface=dirt as that is what it was when they were mapping it)
But anyway, my intention was to try to clarify what you want, by seeing how it differs from combination of existing tags, as I though you posted a thread in order to solicit opinions on your suggestion, in order to produce a better and more widely supported tag.
And not to “force my opinion” as seems that it was understood.
But getting that result will require some work, starting with clearly explaining (ideally by examples with pictures, accompanied by words) how current tagging is insufficient - e.g. something like:
both picture A and picture B here are highway=track + surface=sand + tracktype=grade5 + mtb:scale=2 but A is good (or less technical or whatever) because of reason XXXX, and B isn’t because of YYYY, and I propose to tag A with looseness=XXXX and B with looseness=YYYY to differentiate them
Several such examples, some rounds of back-and-forth, other people adding comments, and it would be a nice proposal with tag useful (and thus being supported in data consumers) by many.
But if you don’t want to engage in such work, sure, you can just start using any tags you like. Far it be from me to try to stop you! But you should been that warned that without the effort to clearly define it before using it, it is quite likely that other people and apps/websites won’t actually use it (or even worse, will use it to mean something different than what you intended).