Do you means, the pdf shows how your page looks like and the html is what you saved and passed back to server side?
If you view your html file you can see the covered layer(colored layer) hadn't been capture succeed.
The problem is we don't have experience on the library html2canvas so we cannot give you better suggestion about how to use it with our mve edition.
About how to solve that, I think you can do something like this:
1. "inspect" the colored layer and make sure it's supported by the html2canvas.
2. Find more information about how to use html2canvas and openlayers.
3. Try some other library to see whether that works for your requirement.
Wish that's helpful.