It's good to keep in mind that there are a very broad range of figures. The Japan one was just the highest I could find with a quick search.

I like this study https://link.springer.com/article/10.1186/s12302-019-0279-9 not so much because they give a definitive answer, but the provide a much better sense of the nuance that bold claims miss. It's too easy to make a bold claim of a number that seemingly contradicts another similarly bold claim. The nuanced approach can often reveal that both bold claims are, in fact, true but not meaningful because they lose significant context.

For example, a lot of reports on water use neglect locality of the use. What the term 'use' means (how much water does a hydroelectric dam use, is that the same sense of use as irrigation?), is there scarcity where it is used? Is it the same class of water as the water in demand (potable / brine / etc.)

The haphazard use of terms has resulted in an insane range of claims of water use per AI query (or lithium mined, or tomatoes grown). The lack of faith leads people to assume one party is lying, but often all of the numbers are accurate in a kind of way. Just not comparable and sometimes not even meaningful

I see you still don't say microplastics are rare. Violently agreeing with each-other, it seems. ;)

Synthetic textiles (clothes, upholstery, carpet, dryer exhaust, washer drainage) are of course the biggest culprits, with most of that trapped indoors with us, or co-located with human activity. If you have a dog that may change the mass fraction, but the MP exposure remains the same (or worse due to additional wear).

Road and tire wear is the other big contributor, again co-localized with population density. That's one of those nuanced cases, because a large fraction of the tire mass is actually natural rubber. The synthetic additives make it categorized as 100% plastic, but this may not accurately reflect reality in terms of the chemistry or hazard-based analysis.