Channel balance options
You're not asking too many questions at all. First, be sure that you need the Series 380 by making sure your bottom attachment matches these measurements:
Once you verify that, you can probably go for a 2830. When you measured the window, part of the weight was resting on the window frame, so your 12 pound measurement is probably pretty light. 2810 is definitely no good, and 2820 isn't ideal. Your "28C" marking tells me "2830" is the best guess, since C is the 3rd letter. Go with 2830 if you're willing to take the risk of guessing. It is your best bet without removing the shutters.
Hm... now I'm not so certain. How did you remove the sash without removing the shutters? Were you carrying the full weight of the sash when you got on the scale? You went with the best method: weighing yourself, then weighing yourself while holding the sash. If you did that while holding the full weight of the sash, then the scale should be accurate. That's odd, though, because 12 pounds doesn't match a "28C" stamp usually. Let us know if you think 12 pounds is accurate or not.
In the meantime, I'll have a balance expert take a look at this. You may want to consider sending photos of your balance attachments for him to see.
Thank you! These photos help a ton. I think our Series 380 balance will work for you. You said that your shoe measures 15/16" deep, but it was probably once 1" and wore down over time due to friction.
If your sash is exactly 12 pounds, then our 2810 may work for you. I would double check the weight just to make sure it's as exact as possible. Like Ben said, your old stamp of 28C corresponds to our 2830 option. This would be too strong for a 12 pound sash, so definitely double check.
It's possible our 1" deep shoe will still fit. We don't have anything more shallow than that, so it's more or less your only option. It's more than possible that your shoe was once 1" deep, but wore down over time due to the friction of use. If ours ends up being too big, you can always shave it down to size a little.