Indeed, Mr. Hailey gives great importance to certain details of the letter shapes as they appear in the CBS reports, while ignoring that the shapes that appear there are almost certainly not the shapes that were present on the originally-faxed document.
In particular, the device that faxed the documents used a pixelization algorithm which preserves relative line thickness at the expense of relative line placement. Small details are always rendered as being a minimum of two pixels high in the vertical direction; the shape of that two-pixel-high image will be affected by the top and bottom side of the detail being faxed.
He notes that Times New Roman has a slanted top on the '1', whereas the samples don't. What he fails to note is that the underside of the top of the '1' is horizontal. Many other letters which appear with deformed serifs have curves on the 'insides' of the serifs even though the outsides are straight. This would again be consistent with the type of faxing artifacts I described.
Proportional spacing is not impossible with a typewriter; someone with a very steady hand could achieve whatever spacing was desired by using one hand to hold the carriage in proper position (with the thumb on the release) while the other hand types. Indeed, many typists have done this on occasion when replacing e.g. a seven-letter word with a ten-letter word (a half-space shift will get one extra letter in neatly, but typing two characters on adjacent half-spaces is ugly); using fudgy spacing is still a big ugly, but not quite so bad.
Of course, the likelihood of a typist doing an entire document that way, with spacing that just so happens to match Times New Roman...
Out of curiosity, did the versions of Times New Roman used by printing houses in 1972 have the modern "f" shape? Look at some old books and you'll notice that the lowercase "f" changed with the advent of phototypesetting.
Animated gif showing overlay of the forgery with text from MS-Word:
Are you talking about "kerning" the f into the fi and ffi?