For super critical flow the effect of the extra energy loss should be visible immediately downstream.
If you only see a change at the very downstream end that seems questionable.
Make your coefficients a little bigger and check if you are getting a difference immediately downstream of each kink. If you are not then you may have a data mistake.
I have not used the coefficients with super critical flow but I would presume they should work.
For 2D flow you can just do a long time series with a constant flow. A bigger issue is that you will have to create a terrain. If you are brand new to this that could be some work.