Implement FV slope limiter for SRHD
Follow up to !108 (merged).
Only adding one ref test using slope_limiter_method = "tvb"
(also should have done that for EulerEq, but nvm).
The results look okish, but not really overwhelming. It might be that one could improve on that by limiting different variables. Atm we follow the suggestion from the Thierfelder paper to reconstruct (limit) the primitives. However, they also use a better reconstruction (WENO).