Typewriters may have been around that could do kerning, but it would have been a semi-manual process, because a typewriter can’t know what letter you just typed, so it can’t figure out the kerning rule.
For those that aren’t familiar, this is what kerning is:
fi
Notice that the ‘i’ actually slides under the top of the ‘f’. Without Kerning, letters that have offset centers look like they have bigger spaces between them. But for kerning to work, you have to deal with characters in pairs. When someone presses the ‘i’ key on a typewriter, the typewriter has no way of knowing whether the previous letter was an ‘f’ which would require the i to be set back, or a Z, which wouldn’t. So if a typewriter supported Kerning, I imagine it would be some sort of mechanical mechanism where the operators can press a key to shift the letter back a bit.
A later typewriter could do it, because it would actually have a small memory to store previous characters. I remember a typewriter in t he 80’s that actually had an 80 character LCD screen. When you typed a sentence it only went to the screen, and then when you hit carriage return or word wrapped, the line would be printed out with a daisy wheel ball. A typewriter like that could do kerning.
So let’s assume that you could kern manually. Why would the author have done so? This was a ‘memo to file’ - a personal memo. Why in hell would you take time to kern letters, superscript the ‘th’, etc? All that is easy now, but it was a pain in the butt back then if it was possible at all. Doesn’t mean it’s impossible, but it’s another data point in favor of ‘forgery’.
So far, for this document to be accurate I think we have to assume that a Texas ANG office just happened to have a very expensive high-end typesetting machine or a high-end typewriter with typesetting features, and that the author took care to use all the typesetting features on a routine memo. Doesn’t seem likely to me.