I think I have a problem with the P2 profile also. In BW's earlier writings about temperature profiles, he suggested a ramp up to 300* and then a hold there to allow most of the water to evaporate from the beans. After that a ramp to the 350-370* range with a hold at that point to allow good caramelization of the sugars. From there an increase in temperature to bring in the first crack and then a further increase in temp for the start of second crack if wanted. If the ramp in temperature is too fast, it favors the formation of chlorogenic acids that cause bitterness, instead of the caramelization of sugars that gives a sweeter character to the coffee.
My rough calculation of a 1# roast using the P2 profile shows a rise in temperature from ambient to 482* in 4 minutes. Then a hold at 482* for 7 minutes, followed by a decrease in temperature to 356* taking approx. 1 minute. The roast then holds at 356* for 5 minutes before ramping back up to 482* in 1 minutes time, and holding there until completion. It seems to me that this profile would lead to 1st crack starting at around the 10-11 minute range and then the roast stalling for 5-6 minutes when the temperature is backed down to 356*. When it goes back up to 482* for the finnish of the roast the 2nd crack probably comes in pretty fast.
I have been using the P3 profile with very good success when anticipating when the start of 2nd crack will begin, and hitting the manual cooling 10-15 sec. before this point. I sometimes will get a snap or two of 2nd crack in the first 30 sec. of cooling.
I actually think that P5 most resembles the guidelines that BW has referred to but the center leg of the roast seems to be a little long. I am going to experiment with shortening this leg by adjusting the profile with time additions and subtractions.
kelppaddy