I think the g/kg unit makes sense. I am more reluctant about % r.H. To me this is still %, nothing else.
If you want this to mean relative humidity, it should be called that way as channel name and item name, not by tweaking the unit.
another channel would be called absolute humidity and have a unit hint of g/kg. (I hope that PR gets merged eventually)
I would not have the UOM system convert between them, as it is only an approximative conversion. If you want such a conversion, it could be done in the binding, between the channels, with the clear disclaimer that it is only valid at 20°C. Another binding may be linked to a device that gives a more accurate value for both absolute and relative humidity, and keeping it in two distinct channels therefore feels more natural to me.

I think the g/kg unit makes sense. I am more reluctant about % r.H.

Honestly if we choose not to do relative humidity then absolute humidity "by mass ratio" should probably also be cancelled. There are in fact three humidity measures -- namely relative humidity (%rH), absolute humidity "by mass ratio" (g/Kg), and absolute humidity "by volume" (g/m³). You don't like the former. The latter is already a standard unit of density. So if we drop those two, we may as well drop them all.

I would not have the UOM system convert between them, as it is only an approximative conversion.

Yeah. I only really did that because I could not find a way to PREVENT conversion (i.e. make it throw an exception). Given that I could not prevent conversion I thought it best to at least return something that more-or-less correct -- rather than absolutely wrong..

@andrewfg I don’t agree with this.
First, relative humidity is a fraction, typically expressed in %. Whether you put r.H. behind it does not make a difference. It is a percentage. It would be nice to be able to put it behind the text, but it does not make a difference in meaning.
Second, converting between incompatible measurements, even in the same unit, is an error. For absolute humidity, while g/kg is a customary (dimensionless) unit, expressing it as kg/kg or % is valid as well. 1% of absolute humidity would be 10 g/kg. I agree adding g/kg as a customary unit because it otherwise is not supported for Number:Dimensionless, and it is a valid dimensionless unit.
If you think about it, also dB is a dimensionless unit. Everyone would agree it does not make sense to convert certain dimensionless measurements to dB, just like you would not convert number of apples to number of pears because it is the same dimensionless unit of ONE.

So creating an almost arbitrary conversion between incompatible measurements is a total no go to me.

see also:
https://en.wikipedia.org/wiki/Humidity?wprov=sfti1

Closing this.