current effective exponents implementation isn't flexible enough

Created by: wilsonmr

CC: @scarlehoff

I helped implement the plotting of effective exponents and next effective exponents in validphys and I was just looking but at the moment it won't be very helpful because you can't specify a different basis than the fit basis.. So this needs to be added, it shouldn't be too much effort

Also as a side note the flavour basis is available in validphys.. line 309 of pdfbases.py and we use it in the fit comparison reports, so I don't think there is an issue with stability of this