I was a bit worried as it seemed this thread started to shift toward elitism... but self corrected. Marshall is absolutely aware and open to the limitations of his experiments. The careful wording is there - but perhaps he could be more explicit with his findings... Certainly no one should translate single data point findings with serious and repeatable science. Marshall doesn't claim this level of authority. Still 'mythbusting' many of the old doctrines of homebrewing are informative and valuable.
I welcome this kind of challenging and open experimentation. Even when the findings are (most often) not definitive, one can take away something. At the least, the debate here has been productive.