In the uniform moving grid with magnetic field test, we have three tests:
subalfvenic and subsonic: everything works fine
subalfvenic but supersonic: also things seem fine
superalfvenic: ripples form. Delayed by using Powell source terms, not prevented.

I would draw your attention to the plots/divB folder first. In the third timestep, the plots look identical, except as if the divergence has been scaled up in the superalfvenic case. The other plots are of other quantities in the second and third timestep. The Bx looks slightly different already in the second timestep (by ~0.1%), unsure if this points to some error or is within calculation error limits.