> this may be normal for BC107?
Read the darn data-sheet! Some geek labored for a week to write it up.
http://newton.ex.ac.uk/teaching/CDHW/Electronics2/DataSheets/BC107_108_109_4.pdf
BC107 110-450
BC108 110-800
BC109 200-800
We infer that they come out of the oven anywhere from 100 to 1000, though mostly far inside that spread. The '07 bin gets the low-end, the '09 bin gets the high end, and when you order '08 you get whichever they have an excess of.
> Most BC107's were at 200-300hfe...with about 12 of 30 @ 350-450hfe. None over 500
Sounds like they have the oven tuned so it doesn't give many ~~110 parts, mostly over 200. And technically Hfe=451 is "reject" for the '07 bin, should be in the '09 bin. (If your meter found some Hfe=470 as '07, you could re-check at the specific test-conditions in the spec-sheet and see if they come out Hfe=449 or if you got a "reject". But you won't get money back unless you have thousands of mis-marked parts. And in fact the high-Hfe parts are more valuable for most audio uses.)
Note that if they finessed the oven to only yield 200-450, all three part-numbers could be sold from the same bin. Incidentally, there would be zero Hfe>600 parts... there is no promise you get any parts at the high end of the spread.
> Are there any real benefits to using transistors with metal can packages?
They look nice.
Someone may argue with me, but I thought it had been shown that epoxy is better in every way, including ways that just don't matter in shirt-sleeve studio work. (We used to think the metal gave a better seal, for oil-well, arctic, and rocket users.)