Whorfianism is certainly oversold, but it's hard to come down completely against it in the face of evidence from color identification.

If it weren't that people make these claims about Pirahã all the time, "it's slightly slower to compare counts of things if you don't have basic numerals for them" is exactly the sort of thing at about the right level of automaticity to sound plausible.