I decided to run a little test on my BCD436HP. It appears the order in which you enter conventional frequencies within the same department has no impact on scan-speed.
I programmed in 189 frequencies with an even distribution of channels in VHF-low, VHF-high, UHF, and 800MHz bands, I created one department of these freqs. in a random/scattered order, and another department with them in perfect ascending order. I timed the duration spent running through each department, with antenna off (so as not to get any delay from outside signal influences), and measured consistently ~2.3 seconds to step through each scenario.. for a scan-rate of ~82 channels a second.. slightly better than the advertised 80 chan/sec. Granted - I was relying on my reaction time for the start/stop of the timer, but with several iterations I achieved results within +/- 0.1 sec. Close enough for government work. It appears the radio (or Sentinel?) does its own internal sort for optimal scan-rate.
I wonder what extent of optimization occurs when you start mixing in other departments of conventional freqs. on the same system, within the same FL, or amongst multiple FLs.. I may do some additional testing later when I've more time..
One of my favorite off-hours scanning techniques is to scan trunking-system repeater-inputs.. this alerts me to activity nearby. . naturally in that scenario scanning-speed is your friend.
I'm pretty happy with the outcome of the first test!