GW/diffusion density calculation should use virtual temperature

During the review of the CAM gravity waves scheme(s) CCPP-ization, it was found that an equation for the air density was using the full air pressure, but the dry air gas constant and regular temperature, which is technically a physical inconsistency.

The original discussion can be found here:
https://github.com/ESCOMP/atmospheric_physics/pull/292#discussion_r2403226331

The response from @JulioTBacmeister was:

>Yes...yes...GW codes are sloppy about water vapor. I guess the thinking is where gravity waves are important - winder mid/high latitudes and in upper troposphere and higher - water vapor is not a big component.

>I'd prefer to leave it as is or even better have a virtual effects flag for GW. Martina and I have grand plans to revamp the entire code in the next year or two. We will address it then for sure.

Thus this issue is here to help everyone remember that in the future the temperature being used to calculate the air density should be the virtual temperature, not the regular air temperature (unless the density is explicitly for dry air). The same potentially holds for vertical diffusion as well.

Also pinging: @jimmielin @mbramberger @PeterHjortLauritzen @adamrher