sigh - read the article again. Download rawdigger. Check the values in the quarry face, not the sky (this is where the ETTR was done) - the sony file is touching clipping on one green pixel, none of the other channels are clipping. The Canon file is 98% to clipping on one green channel. That's about a 1/10 of a stop difference. Please, please read the article and if you'd like to ask a question fine but don't jump to conclusions (e.g. that the sky was the ETTR point, that the difference between clipping and no clipping is significant (it could be a single unit which is .01 of a percent).