ctmixtures: Corrected Moran mutation rate

Calibrated the infinite alleles mutation rate (for all locus neutral copying) against Ewens 2004. Single locus copying yields more alleles than expected for the Moran process in the population, since copying events are necessary to weed out variants to the expected level. This model is suitable for single-locus copying model comparisons, which is sufficient for the current experiment. Needs a "all locus" interpretation of frequency-dependent biased copying, but there are difficulties in interpreting this (see lab notebook).