That seems to be the case with at least one of the IDers I was referring to - every thing in the Gnaphalieae tribe is Anaphalis margaritacea, never mind that California alone has 92 species in that group.
I just tried @anon83178471’s trick on their stats, and it’s 87% supporting IDs, which is actually less than I expected.
Going through State of Matter Life, I have learned some things about what not to observe.
Don’t bother observing plant diseases. You’ll get no agreement about whether it is viral, fungal, nutrient deficiency, or just normal growth, and you’ll be stuck at a three-way kingdom disagreement forever. Same for galls, just replace fungal with arthropod.
Don’t bother observing microscopic algae. You’ll get no agreement as to whether they are Green Algae, Yellow-green Algae, Desmid, Diatom, and/or Cyanobacteria, and you’ll be stuck forever at a three-or-four kingdom disagreement.
Stick to observations like Red Algae that are brown in color; or plant tubers that look like fungi. Then, at least the two-way kingdom disagreement may eventually be resolved.
That’s sad to hear, users can reupload the observation and tag observer with knowledge to what they’re iding, if disagreement has no base in it, it’s the fastest way to deal with it.
Mine splits neatly into thirds 
But interesting to see the split for identifiers I respect, where I can interpret their workflow and chosen use of iNat.
Interesting search. I’m still a little unclear about what is inferred by quantities of the different categories - leading vs supporting vs improving even if it is explained here https://www.inaturalist.org/pages/help#identification.
I understand maverick and check that from time to time to keep mine at bay. The URL you gave will allow checking just mavericks without username which can be an interesting browse.
You did tone your post by saying it is not foolproof and I would agree - I think one needs to weigh things out a little before just looking at percentages and making assumptions. I was looking at a very prolific identifier and even though they had 1.2% leading, that was about 5,460 more leading identifications than I have made. Their total improving (at ~2%) was getting close to 20,000.
Absolutely.
I’ve had, I think 3 in total, brand new accounts (one day old, 2 days old, very new) accounts suggest almost 1k identifications on my observations and then the account gets suspended or deleted, which I always have assumed is because they were just aimlessly hitting agree and got caught doing so. I think some people really do abuse the system and just agree aimlessly, perhaps just for the numbers. I don’t really know. The link can help tell people apart though, because some people agree a lot but are doing so because they genuinely are looking at the observation and they agree.
I think more than anything it can highlight one’s patterns in identifying, which isn’t meant to be an inherently bad thing. For example, my nephew is on iNaturalist and last time I checked his IDs, it was probably 98% supporting, so basically all of them were supporting and very few were leading. He’s very smart and very bright and very good at identifying things but I know how he uses iNaturalist and he really only will agree with things. That’s just the system he uses. It doesn’t mean he’s unknowledgeable about what he’s agreeing with, so I don’t mean to imply that the little chart will indicate if somebody is knowledgeable or not. I think it can just give some insight to some people’s patterns or methods of identifying.
I didn’t expect so many of my IDs to be Supporting. I was expecting more Leading or Improving IDs.
https://www.inaturalist.org/identifications?current=&for=&category=&taxon_id=&user_id=that_bug_guy
I observe tons of plant diseases and galls, and don’t usually run into those issues with it at all. In fact, the galls usually get IDed all the way to species which is nice. But you have to be able to take decent pictures and ID the host plant in the description (and adding it to one of the gall projects helps a lot as well).
Plant diseases can be more difficult because few people actually can get the level of detail needed to ID them. Many of them need a few microscope shots, and some are just unidentifiable without DNA analysis, period. But you can still get some good IDs for a lot of them with a couple of really clear images and a host plant ID.
I find the easiest ways are (if you want Unknowns and Life) to use the “Unknown” iconic taxon in Identify, or (if you want Unknowns but not Life) to add &identified=false to the URL.
You gotta realize you’re not looking at a representative pool of observations when you’re going through state of matter life. There are strong biases in what ends up there, and it’s going to be over represented by hardest to ID stuff and the stuff with the lowest picture quality which makes a kingdom id ambiguous.
I upload galls all the time and get IDs to the order on most of them.
Although plant diseases and galls often end up “unknown” and will be that forever, I disagree that this means we should post them. At least for North America, galls often do get ID’d to genus or species. Most diseases don’t, but I think it’s worthwhile for the few that do. Personally, I have no clue what these diseases are but once in a while I strike it lucky.
For me, what I keep in mind is that the “person” I am supporting is the computer vision. Computer vision works exceptionally well on common California plants, it’s right like 90% of the time. So if a lot of observations have their fist ID given by computer vision, when I make the observation RG, I’m supporting, even though I’m technically the only human identifier on the observation.
It depends on how you “work”, there’re certainly legitimate ways to get most of ids being supporting, but if you want to go out of it, you definitely can if you choose different filters for id.
Mine is nearly half Improving!
https://www.inaturalist.org/identifications?current=&for=&category=&taxon_id=&user_id=fluffyinca
You also never know who might join iNat and start identifying something. For years many hemipteran observations stayed at coarse ID levels until @wongun joined iNat and went through them (since then many other experts have joined, but he was memorably the first in my recollection). And many of my plant observations from Hawaii were incorrect or coarsely IDed until @kevinfaccenda started identifying Hawaiian plants (thanks Kevin!). Just because something might languish now doesn’t mean it will stay that way.
I filter mostly by ‘needs ID’.
Well, it’s the default setting?
Did not the great naturalist Yogi Bear once say “It ain’t over till it’s over.” ;)
So how would I change the ratio of Supporting to Leading or Improving?
In id tab you can choose which taxa you id, at what level, improving and leading ids are mostly done when you id something that is higher than species at the moment.