This shows a problem in the testing. Ambiant (air) temperature can not be read with a 'gun'. They can only read surface temperatures. Even there, on the more expensive ones can be set to compensate for the surface reflectivity.
That doesn't mean the test is invalid. Just their summary interpretation of what they were measuring is incorrect. A surface near the header very well could have been 200* cooler.
It's unfortunate there's no specifics about the coating.
It could have been whatever it is these folks use.
Xtreme Performance Heat Coatings
or one of these products of the same name.
Xtreme Temperature Coating - High Temperature Paint - Header Paint - Stove Paint
Really no information about the products in either case.