Let's be clear here. Nothing outside of some extremely filtered high end equipment is producing exactly and only 222nm light, and no cancer study is isolating exactly and only 222nm light. Except in extreme cases all light sources have some bandwidth. Most UV generators produce tens to hundreds of nm of bandwidth. A common MIG or TIG welder produces plenty of 222nm light in a broad distribution* of UV. UV lamps for curing applications have used arcs to generate deep UV for ages. Lack of a particular study for a particular 1nm band is not a meaningful signal.
We do know a lot about how dangerous different frequencies are, and when you combine the "risk to humans" curve with the "emissions by wavelength" curve you get very low risk.
I do still think we should run additional experiments here before rolling this out broadly for hours-a-day usage, but from what we know so far it looks very good.
Perhaps 222nm is absorbed primarily by dead skin cells - but why is skin the only tissue mentioned? It says nothing about damage to corneal and conjunctival tissues. Eye damage is a major hazard for humans with UV light in particular. "Very low risk" simply doesn't track. "Reduced risk" might.
Eyes, like skin, have a layer of dead cells, and sufficiently short wavelengths can't get through. If you shine enough 222-nm into eyes they'll get dry and uncomfortable, but we're talking about much lower levels.
*Source: https://www.ehime-iinet.or.jp/ehime_e/corp/toyo/ronbun/image...