Interesting article, but I'm curious to know why some lens designs AF more accurately than others. For example the 200 f2 has a superb reputation for accurate AF, but the 135 f2 is notoriously bad. I'm not referring to individual lenses front or back focusing.
The 50 f1.4 has reputation for hunting focus, but the 40 f2.8 is pretty precise.
Any explanations ?
I'm curious about what data you have to show that the 135mmL is nortorius for inaccurate autofocus or that the 50mm f/1.4 hunts.
I've had two 135mmL's which were very repeatable, and three 50mm f/1.4's. One did hunt when I bought it used, I sent ti to Canon and they found broken parts inside and fixed it. After that it was perfect.
Here is the conclusion of a photozone review of the 135mmL, and it reflects what I've seen.
"The Canon EF 135mm f/2 USM L is a highly desirable lens with excellent performance figures in most aspects. It is very sharp and contrasty straight from f/2 and it's outstanding at f/5.6. The bokeh is exceptionally smooth and buttery at f/2 and f/2.8 although it suffers a bit from bokeh fringing at these large aperture settings. Neither distortions nor lateral CAs are field relevant. The only real weakness is the amount of (visible) vignetting at f/2. Flare is generally no big issue but you should use the supplied lens hood in critical situations. The build quality of the lens is excellent and the AF performance is on a very high level. Regarding all the goodness that we've seen from this lens we have to praise it with our rare HIGHLY RECOMMENDED! "