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

[Koen G]

We don't know that™

We don't know anything.  

But in this case we have one bit of direct evidence (the paint over folio number on f42r) and lots of indirect evidence. Such as the fact that may plants had their leaves painted in alternating colors, in a non-naturalistic way. Or the brilliant red leaves of f16v. Or paint obscuring details that the Author should have cared about, like the leaf outlines of You are not allowed to view links. Register or Login to view. (making each 5-leaf bundle look like a single leaf with 5 points), the tail and two legs of the "otter" on f79v, the label on the jar of f102v1...

Whereas the opposite theory -- that the painting follows the Author's intent -- does not have any supporting evidence. Has it?

All the best, --stolfi

I still think a phrasing like "(some of) the colors may not be original" better describes our current state of knowledge. 

Also, I wonder how one could prove that all or some colors are original. We can all agree that the colors were added after the drawings were made. That's usually how artwork is done. But how could one prove that x or y color was applied by one of the original makers?

As said before, the apparent mixing of ink and colorants in one water damaged spot is not evidence, it's an observation that elements of the paints can be seen on the surface where the number was written. This can be because the paint was later there, but also because of the mechanical movement of the pen, the properties of the writing ink, or even the water spill that took place there.