An assertion such as this is best supported with an example or two.
I mean, nobody can guarantee you anything, and it’s likely that out there in Donbas there’s a still signed cycleway that leads through a minefield. Just, I wish that when we discuss tagging principles, we use situations that are reasonably common, rather than hypothetical edge cases.
I mean, nobody can guarantee you anything, and it’s likely that out there in Donbas there’s a still signed cycleway that leads through a minefield. Just, I wish that when we discuss tagging principles, we use situations that are reasonably common, rather than hypothetical edge cases.
What I wrote and what I believe is common practice is that foot=designated or bicycle=designated are about legal designations with no physical implications.
https://wiki.openstreetmap.org/wiki/Tag:foot%3Ddesignated
https://wiki.openstreetmap.org/wiki/Tag:access%3Ddesignated
Designation does not make any statement about whether the way is actually usable for the designated purpose. A rundown designated cycle track is still a designated cycle track, even if it is unusable due to for example smoothness=horrible.
So I searched a little myself, and found a very interesting collection of Real Bad™ examples at:
It features some pretty badly rundown cycleways and footways such as:
cycleway or bicycle=designated or whatever you prefer. But I also would not mind if a mapper decides to downgrade them to disused: or highway=path or smoothness=horrible…
But then, what? My point is: repeatedly quoting edge cases such as those is not helpful. Hard cases make bad law, and, …
…is just unhelpfully hyperbolic. Legal designations do have physical implications. A designated cycleway is implied to be useful by bicycles; the implication is not a guarantee, and there are a few that are actually not. And the Wiki also says as much:
Tag:access=designated:
The route is probably suitable for use by the mode of transport in question (but the surface may have deteriorated after signs were put in place, for example).
I agree with what @Duja wrote above. We have smoothness=horrible for horrible cycleways, exactly as most cycleways are tacitly assumed to be at least =intermediate or better. It is vacuously true that while no legal category has any guarantees for physical fitness, they often do imply some.
In Finland, our strange FtR laws also make the foot/bicycle=yes tags completely bereft of information on paths. Hence, although =designated doesn’t add any legal rights to paths here, they do imply some aspects of usability, and in particular do effect winter maintenance (again, at least in the exotic North). So I’d say that =designated can be interpreted in some contexts to be somewhat ‘between’ purely legal and purely descriptive tags.
…is just unhelpfully hyperbolic. Legal designations do have physical implications. A designated cycleway is implied to be useful by bicycles;
when years ago the city of Rome decided they needed more cycleways, a principal one was signed along the river traversing the center, it had blue cycleway signs, white cycleway pictogram road markings and delimitation markings and was paved with uneven cobblestones for longer stretches and quite some time, (some years ago they decided to lay asphalt on top and now it is generally smooth, https://www.ilcorrieredellacitta.com/ultime-notizie/roma-asfalto-sui-sampietrini-del-lungotevere-per-fare-la-ciclabile-tutti-in-rivolta-contro-la-raggi.html )
The following video popped up conveniently in the suggestions:
illustrating what can be expected on singletracks for MTBs. This one has a happy ending, though.
Ahem.
I do wonder if we’ll ever arrive at a definitive answer as to which highway=* value correctly and properly applies to this 250-metre stretch:
…or if we’ll have to settle for highway=road, the car-centric analogue to highway=path. 
I’ve always viewed the top-level highway=trunk/…/residential tags as a tool for implementing the brilliant rainbow-colored roads of a typical European map. How else to explain the cornucopia of classifications?
(This diagram conveniently ignores the classification-based line weights described earlier. OSM Americana actually overdoes it for the purpose of providing feedback to mappers. Typical American maps make the line weights much more subtle and vary the label typography instead.)
If OSM had been the brainchild of an American and dominated by American mappers familiar with American cartography, maybe we wouldn’t have wound up with so many highway=* values. Maybe we wouldn’t have even called it highway=*. Instead, I think we would’ve had so many endless debates about whether it’s more important to classify freeway/expressway/road/street/driveway versus arterial/collector/local/service that we would’ve all but had to compromise on separate roadway:construction=* and roadway:function=* keys. For sure, we definitely wouldn’t have lumped recreational paths and trails into the same top-level key as streets and highways.
That was a fun thought experiment. 
It expresses no such thing. What it expresses is an identity, a real-world “preset”, if you will. There is no fundamental rule in the English language that a “footway” must come with a certain sign, and so far the stubborn Americans and their silly sidewalks have succeeded at preventing it from becoming a fundamental rule of OSM’s ontology too.
highway=footway may suggest or imply the existence of a certain sign, depending on context. In some cultures and jurisdictions, traffic signs are so consistent and informative that humans and robots alike can orient their entire lives around the prescriptions of traffic signs. The community has chosen to view signs and reality as one and the same. Everywhere else, we necessarily supplement the signs with common sense, intuition, and sometimes a bit of guesswork. Which is why I agree wholeheartedly with:
“Yield triangles” are ubiquitous among North America’s shared use paths and backcountry trails, but they have nothing to do with access. They merely establish the right-of-way, akin to priority to the right. Who is the guest, and to whom shall we be rude?
Maybe something along the lines of priority=wheelchair;horse;foot;bicycle would better communicate the 3+-tiered hierarchy of users without impinging on legal access or suitability.
So you would describe your trail triangle as priority=horse;foot;bicycle meaning: horse has highest priority, then comes pedestrians, and bicycles have to yield to everyone?
You have triangles, while others probably have circles. And what if multiple modes of transport share the same priority? Use commas? Like priority=horse;foot,bicycle;motor_vehicle? Could work, but it will be hard to spot errors.
Generally, this could also be used on crossings to sign which way or mode of transport has priority. It’s one of the missing bits for highway=crossing right now. Maybe some more thoughts are needed, but sometimes, the easiest solution is the best.
priority=scissors;paper;rock;... ? 
Oh, the highway= value is easy. It’s highway=motorway because motorway restrictions apply and are signed as such. It’s the ref= that is disputed…
Yes, that can happen too. Commas, sure, why not?
Circles? The mapper on horseback and the mapper on foot will be there all day bickering over their circular reasoning!
