Did Matthew miscount?

Here's one explanation for an alleged numerical discrepancy in Matthew's genealogy:

https://defendinginerrancy.com/bible-solutions/Matthew_1.17.php

However, I'd like to discuss the issue from a different angle. Suppose someone says the obvious explanation is that Matthew made a computational error, and inerrantists are guilty of special pleading when they flail about for face-saving explanations. Let's play along with that alternative for argument's sake and consider how plausible it is.

i) Even from a naturalistic perspective, it's unlikely that Matthew miscounted. This isn't some off-the-cuff computation. Matthew's numerology is carefully worked out. So it would be surprising if Matthew miscounted. This isn't something he dashed off in haste.

ii) But suppose, for argument's sake, he did commit a computational blunder. Let's take it to the next level. What should we expect in that event?

Even if Matthew didn't catch his oversight, some of the initial readers of the first run of Matthew's Gospel would notice the error. We'd expect word to get back to Matthew regarding his embarrassing blunder. And it would make sense for Matthew to issue a corrected edition–if for no other reason than to spare himself the public embarrassment. It's not like the first run of his Gospel had a wide circulation. It had to be informally copied. It's not like modern publishing where there are, say, 5,000-10,000 initial printings, and if that runs out, the publisher issues another batch of printings.

No, I believe the process would be more like Matthew dictates his Gospel to a scribe, then Christians make private copies. The initial distribution is tiny. 

At most, this would result in two different manuscript traditions, where there were copies of the erroneous ur-text along with copies of the corrected edition. But on that scenario, I think it more likely that only a few copies of the erroneous ur-text would be made, so those are less likely to survive. What would survive is the manuscript tradition preserving the corrected edition, which wouldn't contain the apparent numerical blunder.

But that's not what we have. So I think it's implausible to impute a computational error to Matthew. If that happened, I'd expect there to be a different manuscript record. 

iii) Notice that I'm not proposing a solution to the crux. I'm doing something different. I'm pointing out that on closer scrutiny, the naturalistic explanation is implausible. We don't need to know what the right explanation is to eliminate the naturalistic explanation.