It is the performance win for similar looking results that I find improbable. For a box blur to look like gaussian blur, you would need multiple passes. Even though each pass is now O(1) instead of O(n) (with n the blur radius), due to caching effects I think a gaussian kernel would still be faster, especially for the small blur radius as described in the article.
That link is not a box filter, as it still uses weights to approximate a gaussian convolution kernel. It just uses some special hardware to do less texture fetches. But that is a constant 2x improvement over the full 1D convolution, not the box filter O(1) approach that the article suggests that browsers are using.
You've moved me a place of uncertainty here. I had some confirmation that _some_ browsers use box blurs for this effect, I _know_ some game engines use multiple box blurs to approximate Gaussian blur (having seen the code myself).
I updated a few sentences in the article to reflect that uncertainty. Thanks!