Add example of hybrid CG-DG discretization for solving NS

[lindsayad]

Reason

CG pressure and DG velocity (at least with monomials) naturally satisfy LBB stability and so are naturally attractive from an implementation standpoint as they do not require PSPG like equal order CG nor Rhie-Chow like collocated FV.

Design

Try to use as much existing code as possible, but add kernels (particularly DG kernels) where needed

Impact

An attractive discretization option for people like @joe61vette and for tossing at our upcoming assessment suite for TH problems

[lindsayad]

This refs the deficiencies (thus far) in a single-phase Rhie-Chow implementation discussed in #22637. Note that there are modifications to Rhie-Chow for two phase which are also discussed in that issue.