Great theories. Here is the exact format I wrote it in, but I changed all the characters. I do remember maybe converting it to hex or binary, or at least considering it.
0486 | 4AE5 | 41BC
A58F | 3FBA | 1AD7
34F6 | 2AEB
37A4 | 4AAF
7?A4 | ?42? | 6F?? | ??D8
71?D | 6?A? | ?FCA | 86B?
32?F
7F?B
Are the question marks included inside the backup you made, or are they part of the redaction?
I wish you the best of luck to successfully restore your wallet. Given the fact that it's "only" 14 characters missing there is a good chance at some point within the foreseeable future it will be crackable by a strong (cloud)-computing-network within a reasonable timeframe.
https://www.proxynova.com/tools/brute-force-calculator/ currently gives an estimate of ~1000 years, but I think that's based on the time it takes an average home computer to crack this. Meaning a powerful super-computer could quite likely already crack it within a few days or weeks.
With the current rate of Bitcoin this is most likely still not worth it, but if Bitcoin climbs up to $1M/BTC and computational power is getting cheaper and cheaper in the meantime, it will most likely be a profitable thing to crack.
Only 14 hex characters missing? That should not be too hard to brute force since there are only 16 possible hex haracters for each position, and if you know how to program in CUDA it should be simple to write a program that makes all of these permutations and then creates the compressed or uncompressed private key and its address from which you can check for balances against, using publicly available lists.