Coming late to the discussions, but my responses:
1) I think having individual pop literacy is a great step forward - I could never understand how literacy could instantly change when a province was incorporated in a new country. This means that adding new (less literate) provinces with your national pops will hurt your RP generation.
2) Explicitly modelling bureaucracy and hence capacity to carry out functions of a modern state is a great enhancement. Presumably the literacy of bureaucrats will affect their performance.
3) Including government expenditure back into into incomes of bureaucrats and clergy is a neat concept. Presumably also reflected in officer/soldier incomes?
4) Will there still be a shift in clergy impact on education/raising literacy with advent of Darwinism as for V1? If so, perhaps you could implement the suggestion of altering the allocation of education impact between clergy and clerks via decision (with some choices only being available with requisite tech):
a) Clerical education (Clergy high weight)
b) Multi-model education (both Clergy and Clerks have weight)
c) Secular only (Clerks high, Clergy low)
Alternatively, or even better, you could make the shift be from Clergy to Bureaucrats.
Not sure how you would interact this with income flow from education slider - perhaps cause some shift from clergy towards bureaucrats (or clerks) as education model changes.
This would enable country wanting to go the low con route to maintain Clergy in teaching positions, whereas another nation could go higher con by using bureaucrats.
5) Now there is pop literacy, will there be limits of conversion of pop type based on literacy, e.g. labourer to craftsmen only if lit >x%; to bureaucrat only if >y%, etc? If their literacy reduces (eg due to underfunding education) will pops in professions requiring high literacy (eg bureaucrats) devolve?
6) How will pop merging/splitting work for literacy? Weighted average of old pop literacies? Could you split a pop into differential literacy groups (e.g. to convert high lit into clerks/bureaucrats and leave low lit as labourers etc)?
7) Perhaps bureaucrats should have con raising effect - ie the more intrusive the government is in people's lives the more they care about whether it is done right (ie a way they are happy with). Hence to keep people at low con, you want lots of clergy and relatively few bureaucrats (or at least only lowly funded bureaucrats who can't interfere too much) = low intrusion, so minimal attention to what government is doing.
8) What about corruption events when you underfund bureaucracy? I guess that is captured in crime building/crime fighting mechanics.
