I think veloclinic gathered power data from some Twitter friends, then ran all kinds of regressions, came up with his own model, and presented the results in a way similar to Coggan (exact same graph layout etc.).
Although I am not 100% sure, because sometimes veloclinic's blog is a bit...