I stand by my verdict. He was stating that ring fenced measurements are blown out of proportion because they do not reflect real life behaviour. I'd expect a citation otherwise I am happy to dismiss it as silly.

Well, in effect what you're arguing is that because something was tested in isolation, you feel justified in generalizing it to a situation it was not tested in.

Absent further evidence, that seems like a safe conclusion given the caveat that it hasn't actually been tested.

Hold on a second. It's not that there is no evidence, I just personally haven't bothered to crawl the internet looking for it. A quick google search found the following: "High- and moderate-GI choices may be improved by pairing them with low-GI foods or healthy sources of lean protein or unsaturated fats. For example, the GI of white bread will be lowered if it is eaten with a bowl of chili or spread with peanut butter." I.e. eat your white rice with some chicken and olive oil and you'll be "okay".

http://clinical.diabetesjournals.org/content/29/4/155.full