I assume the AFM data has a safety margin factored into the 1st principle data to allow for differing reaction times from line pilots compared to test pilots.
The aircraft is certified under FAR25 regardless of the method of takeoff analysis computation, so if FAR25-109 calls for a 2 second reaction time, this has to be accounted for in the certification data, 1st principle data and old style AFM data.
Look at appendix 1 of
http://www.airweb.faa.gov/Regulatory...7?OpenDocument AC25.1581 shows the requirements for an AFM regardless of if its a paper version or electronic version.
You have said that Embraer supply an AFM or FP option, I presume that you are talking about the E135/145 aircraft as there is only one option available for the E170-195 series of aircraft based upon 1st Principle.
If you are finding differences, it because of the calculation methodology, we had the same problem running B744 analysis on a mainframe, 2nd Principle and 1st Principle, its all based upon the same certification requirements and data, but the answers are slightly different.
None of the manufacturers will tell you the exact methodolgy that they use as its considered a trade secret, but considering that Embraer (E-series) have managed to cut 40% off the calculation time by changing the methodology, it demonstrates that there are various metholdogies available.
Mutt