Use eventual future host-side constituent properties feature for aerosol radius (VOLC_RAD_GEOM, DGNUM, ...)

### Issue Type

Code Clean-up

### Issue Description

During the review of https://github.com/ESCOMP/atmospheric_physics/pull/372, @nusbaume originally wrote regarding `VOLC_RAD_GEOM`:
> I don't think it matters for this PR, but long-term I am not sure it is ideal to have the radius of the volcanic aerosols be its own constituent entry, because it is really not a constituent per se (i.e. something that has "substance" or mass), but a property of the volcanic aerosols themselves.
>
> If this is the only chemistry/aerosol scheme you know of in CAM that does this then I think we can ignore this concern, but if multiple aerosol configurations do it then it might be worth opening an issue so we have a place to start thinking about and discussing possible alternatives.

I wrote:
> I think this will matter in MAM too where the geometric-mean diameter of the climate list modes ("DGNUM") is stored as a pbuf field and I built in the assumption that it will also be a constituent in the modal aerosol state port in the bigger BAM/MAM optics PR. I agree it is best to have it as an attached property to each bin, and this slots well with the development of a host-side properties object. I will open an issue for this. Thanks for the heads up!

### Will this change answers?

No

### Will you be implementing this yourself?

Any Software Engineer can do this