- First of all, I'm happy that you use mu=3 GeV for the
   renormaization scale of the charm quark mass. mu=mc is too
   small, as you also write in the document.

- In 4.2 you mention that you use "three-loop MS-on-shell conversion"
   as "highest available perturbative order" but later you refer to
   [1] where 4-loop corrections have been computed.

- In 4.4 you propose to compute mb^pole from mb(mb)=4.18 GeV
   using the 4-loop OS-MSbar conversion from [1]. Note that
   for the bottom quark there is no convergence: the 4-loop term is
   not much smaller than the 3-loop and 2-loop coefficients.
   I assume that the dependence of \sigma(pp->Hbbar) on mb^pole
   is weak and \pm 100 MeV in mb^pole are not relevant.
   Then the procedure is OK.

- Do you also need the top quark MSbar mass? In that case
   the the 4-loop conversion formula behaves well.