What's the evidence that the colors were added later?

[Jorge_Stolfi]

even if the ink is applied over the (water soluble) paint, it will seep through to the vellum and react with the vellum, creating a dark patch under the paint.

To properly answer this question we would need those microscope images.  To see how much pigment there is, how the particles are deposited or embedded into it, etc.

But, until then:

• The green paint is not just the solid pigment particles; it is particles embedded in some binder, that prevents them from falling off the vellum.  Thus, in the scenario above, the ink would have to get through this binder, too.
  
• The dried ink is not just a molecule-thick layer.  It has some thickness, which I expect to be more than the diameter of the green pigment particles.  Thus, if the ink was applied over the paint, the pigment particles would be seen to be embedded into it, not resting on top of it.
  
• To write the folio number, the quill had to rub against the vellum with a minimum of pressure.  If the paint was there already, the pigment particles should have been pushed aside or dragged along, at least in part.  This effect should be visible under the microscope.
  

So, while the scenario above does not seem to be impossible, it seems to be incompatible with the observations by Rene and others...

All the best, --stolfi