You can get UV LED's in different wavelengths, but 253nm is the very effective for disinfection (even though it does suffer from the other UV damage that you mentioned). It's slightly wider wavelength might make it more possible to make using existing LED manufacturing techniques.
This would open up a lot of possible germicidal sanitation technologies from integrated germicidal LED's on door handles, bulbs used with home automation to disinfect a whole home once everyone has left for the day, to Roomba-like devices that disinfect floors.
[Aside, I believe the peak germicidal wavelength is 265nm, 253 is just the nearest line that mercury lamps have].
Some of the applications you mention might be problematic because many indoor materials aren't particularly UV durable. The idea that conventional home automation might drive germicidal lamps is kinda frightening to me. ALEXA DEACTIVATE CANCER TUBES. ALEXA! ALEXA! :P
How hard/possible would it be to make a ~1m tube of 1 -3cm diameter, with a LED strip suspended down the centerline. a charcoal filter and a very light fan on one end, and face-mask on the other end. the fan and LED could be powered by an 18650 dangling from a lanyard/necklace. some kind of valve in the face mask would have to vent your exhales and then toggle for the inhales to draw through the tube.
How far off am I on the parts? You can get a 10Wh 18650, can you run a 1 meter strip of UV LEDs at 5 or 10W? Would that be enough power at the right distance (radius) and for enough time (an inhale would draw air through quickly) to do the job?
https://news.ycombinator.com/item?id=22233032
Link to paper: