dust_emis_fact needs separate values for F and B cases

### Issue Type

Other (please describe below)

### Issue Description

I am documenting our discussion in PR #1509 here as an issue. #1509 supports @cecilehannay 's spinup run in issue #1507 with multiple namelist parameter changes in the `namelist_defaults_cam.xml`. In PR #1509, @cacraigucar requested to change the dust tuning factor `dust_emis_fact` in the 1° or ne30 horizontal resolution from 4.0 to 3.2. I mentioned that 3.2 works better for B cases in 1°, while 4.0 works better for F cases (freeruns) in 1°. But 3.2 was not officially put inside the namelist defaults, and 4.0 has been in the namelist defaults since I did my latest dust tuning in PRs #1423 and #1426 a few months ago. @adamrher and I agreed to use a smaller `dust_emis_fact = 3.2` to allow larger dust emission fluxes for Cecile's B cases.
The aerosol and chemistry communities (and ACOM) mainly work with aerosols in F cases, so I generally recommend maintaining 4.0 as the dust tuning factor (but I know there are people running B cases too). I told @cacraigucar that, if possible, it would be great to have `dust_emis_fact` values separately stored for F cases and B cases in the namelist defaults, but I am not sure if this is technically possible. I am not sure how to settle this, so I create this issue to see if @cacraigucar or anyone else has any proper idea.

The following is a few instances of what was merged in #1509. They were originally 4.0D0.
```suggestion
3.20D0
3.20D0
3.20D0
```

@PeterHjortLauritzen @tilmes

P.S. Dust emission in the B case is lower mainly because CLM%BGC in B cases has generally higher LAI than that in CLM%SP in the F cases, thereby reducing global dust emission by 20–30 %. The land people like Will Wieder and Dave Lawrence are aware of this.

### Will this change answers?

Yes

### Will you be implementing this yourself?

Yes, but I will need some help