Double-check cmflq (total_water_flux_due_to_shallow_convection) standard name

This issue arose during the review for Hack shallow convection CCPPization https://github.com/ESCOMP/atmospheric_physics/pull/195

@nusbaume wrote about the `cmflq` standard name:
> Is this actually "total" water, or just water vapor and cloud liquid water? Technically cloud ice would be included in total_water but I don't think anything is actually done with ice within the hack scheme outside of the general constituent tendencies.

@jimmielin wrote:
> I am not too sure about this. Some parts of the code explicitly state cloud liquid water, so maybe "total water" is intentional, but you are also correct that nothing is done with cloud ice except from convective transport. The units for cmflq are also W m-2 which is unusual so I am not sure if we want to revise the standard name.

Current definition
```
[ cmflq ]
standard_name = total_water_flux_due_to_shallow_convection
units = W m-2
type = real | kind = kind_phys
dimensions = (horizontal_loop_extent, vertical_interface_dimension)
intent = out
```

defined as
```fortran
! The following variables have units of w/m**2
cmfsl (i,k+1) = cmfsl (i,k+1) + fslkp*adjfac
cmfsl (i,k ) = cmfsl (i,k ) + fslkm*adjfac
cmflq (i,k+1) = cmflq (i,k+1) + hlat*fqlkp*adjfac
cmflq (i,k ) = cmflq (i,k ) + hlat*fqlkm*adjfac
```

Tagging @nusbaume and @adamrher for future discussion.

Thanks!