So my understanding from this thread is that through 1st and 2nd while rippin through the gears the m3 is faster or atleast as fast as the m5. After that 3rd and beyond the m5 pulls hard away.
Also to the above poster regarding comparing wheel torque at 3.5k RPMS. You really cannot compare single points at certain single rpms to one another. You have to do an area under the curve in order to accurately represent things. This is not overly hard to do since the graph of the m3 is actually so linear you can simply use normal math of a square, dividing it by 2 to get a triangle compared to linear regression needed to calculate the m5s wavy curve AUC>
Nonetheless what you compare is a real life scenario which is 6k-8400 rpms for the m3 and 5.5-8200 for the m5. These are the useable rpms at full throttle and the rpms do not drop below about 5.5-6 for either of them in full throttle pulls. The only time is for an INSTANT from a standstill but it hits 6k in the blink of an eye. So the AUC of 5.5-8400 is what you would compare to the m5 5.5-8200 AUC.
I am willing to bet the 1st gear AUC is likely almost identical giving the shorter gearing in the m3. Gearing has a huge multiplication effect. Which is why some people do not understand that it is almost always more beneficial to take a car to its redline even if its not making as much engine power. Staying in a lower gear longer multplies torque by so much more than the power you lose being out of your "power curve"
After 2nd I would think the m5 would really have much larger numbers and then from there on out it would be no contest