The evolution of P/F vs T/K raises important questions

[npcompl33t]

Yes look, P cannot be an abbreviation for more than 1 K character, because its length distributions are already extended. So how would abbreviating K words result in longer P words? 

It makes sense if K was the form that was actually abbreviated, and Kch -> ke sometime in 'archaic' voynich, and this change never made it over to P.

   

And to stolfi's earlier obsrvation around lengths being binomial distributions.