hydrostatic equilibrium - Are larger stars rounder?

In terms of mean angular velocity, the distribution of rotation rates among main sequence stars is well known. Allen (1963) compiled data on mass, radius, and equatorial velocity, which was then expanded upon by McNally (1965), who focused on angular velocity and angular momentum. It became clear that angular velocity increases from low rates for spectral types of G and below before rising to a peak around type A stars and then slowly decreasing.

Equatorial velocity continues increasing to mid-B type stars, before slowly decreasing, but because of the increased radii of O and B type main sequence stars, the peak in angular velocity occurs before this. As part of Jean-Louis Tassoul's Stellar Rotation notes, many O type stars have rotational periods similar to that of the G-type stars like the Sun!

The distribution is not smooth and uniform (McNally noticed a strange discontinuity in angular momentum per unit mass right for A0 and A5 stars; see his Figure 2); Barnes (2003) observed two distinct populations in open clusters, consisting of slower rotators (the I sequence) and faster rotators (the C sequence). Stars may migrate from one sequence to another as they evolve. Interestingly enough, stars on the I sequence lose angular momentum $J$ faster than stars on the C sequence:
$$frac{mathrm{d}J}{mathrm{d}t}propto-omega^n,quadtext{where}begin{cases}
n=3text{ on the I sequence}\
n=1text{ on the C sequence}\
end{cases}$$
Here, of course, $omega$ is angular velocity. These results obey Skumanich's law.

Oblateness can be determined from mass, radius, and angular velocity as
$$f=frac{5omega^2R^3}{4GM}$$
Using this and McNally's data, some quick calculations get me the following table:
$$begin{array}{|c|c|}
hline text{Spectral type} & f/f(text{O}5) \
hline text{O}5 & 1\
hline text{B}0 & 1.28\
hline text{B}5 & 1.84\
hline text{A}0 & 1.67\
hline text{A}5 & 1.35\
hline text{F}0 & 0.482\
hline text{F}5 & 0.0387\
hline text{G}0 & 0.000314\
hline
end{array}$$