Support MLE fits to binomial distributions

[hartytp]

• Fitters no longer calculate fit significance - that will be handled by separate validators in future (https://github.com/OxIonics/ionics_fits/issues/110)
• Support binomial MLE fitting

TODO:

• [x] for binomial fits we should disable model rescaling since we don't want to muck around with probability distributions. That will come out of https://github.com/OxIonics/ionics_fits/issues/105 but needs a little though. For now I've just worked around this so we can fix later...
• [x] error bar calculation
• expand testing? @cjbe any ideas? The gold-standard thing to do here would be to generate a bunch of random datasets and then look at the parameter means and variances and compare against the variances the fitter gives us!
• [x] update readme after changes to fit significance calculation (mention validators etc)

[hartytp]

I'm happy to merge this now as a PoC. I've opened separate issues about expanding testing etc.

[hartytp]

I've checked that this provides a good (unbiased) estimate of the parameter values and calculates standard errors correctly (see below). Needs a final proof-read then I'll merge.