Averages being...well..."averages", that sounds way too high. Especially on a longer track, where you wouldn't have the motor loaded as hard or drawing as much current.
To have 45 as the mean over 5 minutes, you must have some huge draws to offset the very much lower to near-zero draws you would expect during the majority of the run.
It sounds suspicious statistically to me when your "average amp draw" is half your pack's rating. Sometimes, one can construct statistical models that discover exactly what one believes it should discover.
BTW -- I'm not totally convinced that I'm correct, just that your 45A average number seems unrealistic to me. Using that same methodology, my Mabuchi motor draws an average of 12A over 5 minutes...does that fit? Does it follow that a 3.5 brushless is only using 3.75 times as much power as a Mabuchi over 5 minutes?