Thanks for the answer, Richard. I do sometimes wonder about the validity of some of the magazines' testing regimes, especially when they quote resolutions with differing lenses. As we all know, some kit lenses are less good than others and the same (probably) applies all the way up the ranges. Same with their lens tests, unless there's a constant factor, comparisons are fairly meaningless unless the test is a comparison of kits.
Do you get loaned review samples or do you pick one from a retailer's stock? I know that in some industries, review samples are sometimes more rigorously QC'ed than random samples... One would hope that all examples of a lens (or body) would be more or les the same but I'm pretty sure some are less good than others.