ENERGY, keep open: p3 F version has clippings without fixers

In micro_p3_interface.F90 Q tendencies are computed via
```
temp(:ncol,:pver) = th(:ncol,:pver)/exner(:ncol,:pver)
ptend%s(:ncol,:pver) = cpair*( temp(:ncol,:pver) - state%t(:ncol,:pver) )/dtime
ptend%q(:ncol,:pver,1) = ( max(0._rtype,qv(:ncol,:pver) ) - state%q(:ncol,:pver,1) )/dtime
ptend%q(:ncol,:pver,ixcldliq) = ( max(0._rtype,cldliq(:ncol,:pver) ) - state%q(:ncol,:pver,ixcldliq) )/dtime
ptend%q(:ncol,:pver,ixnumliq) = ( max(0._rtype,numliq(:ncol,:pver) ) - state%q(:ncol,:pver,ixnumliq) )/dtime
ptend%q(:ncol,:pver,ixrain) = ( max(0._rtype,rain(:ncol,:pver) ) - state%q(:ncol,:pver,ixrain) )/dtime
ptend%q(:ncol,:pver,ixnumrain) = ( max(0._rtype,numrain(:ncol,:pver)) - state%q(:ncol,:pver,ixnumrain) )/dtime
ptend%q(:ncol,:pver,ixcldice) = ( max(0._rtype,ice(:ncol,:pver) ) - state%q(:ncol,:pver,ixcldice) )/dtime
ptend%q(:ncol,:pver,ixnumice) = ( max(0._rtype,numice(:ncol,:pver) ) - state%q(:ncol,:pver,ixnumice) )/dtime
ptend%q(:ncol,:pver,ixcldrim) = ( max(0._rtype,qm(:ncol,:pver) ) - state%q(:ncol,:pver,ixcldrim) )/dtime
ptend%q(:ncol,:pver,ixrimvol) = ( max(0._rtype,rimvol(:ncol,:pver) ) - state%q(:ncol,:pver,ixrimvol) )/dtime
```
Not clear if this code is ever active, but if it is active with significant values clipped (also, unlikely), then it would lead to mass and energy leaks.

I haven't seen such code in p3 cxx version yet, but since the same code is in v3, scream is probably a better space to talk about it.

I would assume simplest mass and energy fixer for water forms would be enough here.