It seems that the first digit falls in the set {0, 5, 3}, where:
0 = 10x
5 = 8x
3 = 12x
I can't disagree with this.
The second digit falls within the set {0,5}, independent of the first digit. What could this mean? (What happened to 1, 2, 3, and 4?)
That's why I think it's a coding not a sequence. And you can't say anything about the independence of the two numbers. We have four observed combiations (Thanks, Bryce for another one).
00, 50, 55, 35.
We haven't seem more variations which I would expect if they were incremental e.g. we've not seen 30 (but we don't have a huge numbe of 12x serials).
The last four digits are clearly sequential within the range {1, 2, .... 9999}.
I agree with this too.
Unless more variation is found with the second digit, therefore, there would appear to be a max of 20,000 produced for each magnification.
Not sure how you get this ... but I can't believe anyone would produce the same number of 12x as 8x.
My suggestion that they are sequential with different "tens of thousand" starting points (i.e. two digit prefixes) and that you can have:
up to 34998 10x but I think less than 10,000 were made (less than 9000?)
up to 14998 12x (not enough info to say anything else but perhaps 1000?)
up to 50000 8x (at least 6,000)
seems to match bin usage a bit more closely but I don't think they ever got to 50,000 8x32 SE!
Actually I suspect the 55xxxx 8x32 are special or different in some way hence the new serial (which would imply the first kind of 8x32 has under 10,000 units). Remanufactuered? A second design run? Different coatings?
My guess is that the second digit represents separate manufacturing runs of up to 10,000 units each. The two runs might or might not differ in some way, such as coatings.
I suspect that this serial scheme doesn't include any batch system that's usually made explicit in modern serials which can include country, factory, year, week, batch number and number in batch (a lot of computing serials are generated this way).
I think we have a set of prefixes (preficies?) and a increasing sequential serial number. And I wouldn't be at all surprised to find these were all made in the late 1990s (97, 98, maybe 99) in a collection of large batches and have since sat in storage or in distributors or in retail.