Other than grandchildren I don't shoot sports but love to shoot the animal kingdom. I'd say about half my "animal kingdom" shots require AI Servo AF.
The m3's high ISO noise performance and 14 bit RAW resolution appeal to me in a big way. I've been shooting a 1DmkII almost from the time it hit the streets and the extra resolution and frame rate of the m3 is a ho-hum to me. Having seen Paul Bestwick's portraits comparing the 1DsII and m3 it is clear to me that the m3's detail extraction will also make it a great landscape camera. I've used my m2/17-40 for landscapes (my other love) and the extra detail afforded by the m3 won't hurt me at all.
I've been an engineer involved in high technology developments all my life and I'm terribly concerned about the problems posted by a limited number of credible photographers; these problems have ranged from AF issues on static as well as dynamic objects. There has been one case of a good m3 camera going unacceptably bad from the standpoint of AF; this "shift" in performance took place over a two week period. Of course, there is also the temperature dependence reported by RG. My take on the "shooter experience" aspect of the m3 issue is different than most folks; seems to me the shooters with zero previous 1-series experience are the ones that report happiness. Most of the unhappy "credible" shooters have been shooting 1-series bodies for quite a few years; in fact they still have 1DII and 1DIIN bodies in their possession and comparing the m3 to their frame(s) of reference is a very easy thing to do.
So, where do I sit? Well, about a month ago I decided to wait six to nine months to see how the m3 plays out. My experience as an engineer listening to many problems just like this (i.e. with mixed apples in the basket) tells me the problem is a design issue that cannot be fixed via firmware (firmware changes can minimize certain aspects of the problem but IMO will never get to the root cause).
So I'm going to wait it out. I have three (had four but lost one) competent bodies and lots of lenses that play perfectly with those bodies and therefore there is no need on my part to rush to what may be a bad judgment. However, I badly want an m3 and if by the end of next week it appears my chances of getting a good m3 will hit 99 percent I'm jumping on the bandwagon BECAUSE I always reserve the right to be wrong and/or change my mind.

Unfortunately, right now, looking at polls conducted at other forums and my own short list of credible shooters, it looks to me that the chances of picking up a good m3 is like a 50/50 coin flip; I don't feel comfortable with those odds. Now odds can be deceiving and one must be flexible when it matters; if I had a 50/50 chance to pick up one of the Coors twins ......., well, you get my point.
Regards,
Joe Kurkjian