Just out of curiosity, how can anyone say it is 100% ANY locality without any collection data or knowledge of it's lineage?
While I agree that it looks to have some Sambava-type qualities, you don't usually see the red pattern along the spine that this one has in Sambavas. Ambanja? No. But this animal does look somewhat similar to one of my old holdbacks that was wc Ambanja x wc supposed to be Ambanja female.
It would be pretty much impossible to have positive locality data, and even some importers get them them mis-labelled... To open a shipment labelled Nosy Be (for example) and see what appears to be anything but, you have to question the exporter. To call them what they're labelled when they're clearly not would be careless. Green tree python owners/breeders have the same problem (and there's lots of commotion between them because of it), but they breed less true than panther chameleons in my opinion, and resemble the other locales more (different locales of chondros have very minor differences, but panther chameleons are more defineable between locales).
There's a few staples that are true 95% of the time with Sambavas, perhaps more than any other locale. Without getting into them, nearly all crosses would have some amount of blue in them, whether they were partially Ambanja, Nosy, Ambilobe etc, it's just a dominant color. There's exceptions to any rule, but the red on yellow with the perfect "U" shaped bar and zero blue makes it pretty clearly a Sambava. Details vary between individuals (red along the spine etc).