The only way I know of to estimate what percentage of bucks are not captured on film is to use a modified version of "Mark-Recapture" statistics. With Mark-Recapture studies, animals are captured, all are marked and then released. Later, animals are recaptured and the percent of the second capture that were marked from the first capture gives you an estimate of what percentage of the total population was captured the first time (generally, these later captures are run over and over to get a good estimate). So if 30% of the animals recaptured are marked, then that first capture caught about 30% of the total population.
I can use the bucks captured on film as the "first capture," and then bucks harvested as the recapture, and the number caught on film previous to harvest as the "marked" deer. Since 2003, of the 63 bucks harvested, all but 2 were "marked" (previously caught on film). So those bucks caught on film each year represent around 97% of the total buck population. That's a pretty good census.