I did a test with the BAR. Set damage to 1, burst to 10000 rounds, accuracy to 50% on all ranges, got a target with 1 target size and 10000hp, made sure no vet is applied.
A single burst round left the target with ~50%.
Then I used to 2 targets, standing next to each other. This time a single burst would not drop the hp by 50%, instead it lost about ~31% while the other model, which was not targeted, lost about ~18%.
The myth of "re-rolls" is hereby debunked. Therefor guns with focus_fire enabled are superiour. The primary target however will be prioritized by weapons without focus_fire. By which modifiers the secondary targets are selected I can't say for sure - not of my interest though.
Could you repeat the same experiment with 5 targets on the same location? I'd be interested to know of the result.
I've only tested this in the past with two PGren squads, one with focus fire set to on, and one with set to off, and the squad with focus fire set to off was winning by a large margin, when it was squad vs squad.
Conversely, when it was 1 model vs 1 model, I didn't notice any difference.
Your testbed is much better than the one I used, so it would be interesting to get hold of more data to know about how this works precisely.
PS: make sure you remove all other variables from your testing; e.g., min-max cooldown, burst, etc etc. The only thing left at random should be accuracy.
Alternatively, if you can't be bothered, let your guys fire for 20 seconds at a time and post the results. Bursts can be variable.