A followup...
I want to report my findings about Varro Tyler's Focus on Forgeries copyright 2000, and his Batum Genuine/Forgery signs on page 27 & page 28. I had the older edition before (1993), and it did not include Batum in that book. I now have the 2000 edition.

Keep in mind, I am not an expert, nor do I play one on SCF.

But I have taken an interest in the "Aloe tree" Batum forgeries, so take this discussion for what it is worth.

• His Forgery first type is my nomenclature Forgery II, and Forgery second type is my Forgery I. Keep that in mind.

I will be using his nomenclature here for consistency for those that have the book.
• He directly compares a 1919 Scott 1 5 kopeck green genuine with a Tyler first type forgery (My Forgery II). As he clearly points out, for the kopeck values, the genuine will have six white dots over the right value tablet, while the first type forgery will have seven dots.
And the Tyler second type forgery (My Forgery I) will also show seven dots.
He does state that the genuine kopecks have a joined KA in the upper inscription tablet. That is indeed true for the plain Scott 1-3 (not overprinted) Kopecks.
But, similarly, does the first type forgery always have a joined KA for the Kopeck values (Scott 1-3)?
Yes, the joined KA is common for the Tyler type one forgery,
but not always.
Take a look at this example....
Batum "Aloe Tree" Scott 1 5k green (genuine)
Tyler Type one Forgery ( My Forgery II)Note the Tyler type one 5k forgery is not joined at the KA.

• Bottom Line: For Kopeck (Scott 1-3) values, the Genuine values appear to have the joined KA, while the Tyler type one usually are joined, but not always, as shown above.
Why does he discuss this? Because the Tyler type two forgeries (Which he does
not illustrate), are never joined at the KA.
So is his "KA" test valuable for the plain (Scott 1-3) issue versus Tyler forgery type one and Tyler forgery type two?
Yes, but there
are exceptions.
But remember some other easily found signs: The Tyler type one forgeries are always found on white paper, and have the easily seen "Curved Branch" sign (See earlier posts).
Tyler then shows the ruble values (page 28) with the 1919 Scott 4 1r red brown and a Tyler type one forgery. (He does not illustrate Tyler forgery two.) His discussion is good for the plain series only (Scott 4-6), as we will see in a moment.
Again, the genuine copies have a joined KA (except if there is a printing flaw, which I happened to note in an example I have).
The ruble values Tyler type one forgeries also have a joined KA. (But recall the 5 Kopeck value above did not have it.)
And, significantly, the ruble values Tyler two forgeries for Scott 4-6 do not have a joined KA.
See below....
1919 Scott 4 1r red brown, Tyler type two forgery, Tyler Type one forgery(Aside- Note the extension of the vertical right outer frame line above the upper horizontal frame line (Right upper corner) of the Tyler type one forgery? Tyler states the type one forgeries always show this sign for the type one forgery ruble values. I checked my stamps, and it appears so, although sometimes the extension is blunted/difficult to see.)
Genuine-KA joined
Tyler type two forgery- KA not joined
Tyler type one forgery-KA joined
So far, so good.
But this "KA" rule breaks down when discussing the "British Occupation" overprinted "Aloe tree" stamps! Look at the initial post by danko, and his overprinted genuine Scott 17 2r salmon pink does not have the KA joined. In fact, when I reviewed all the Scott 13-20 overprinted illustrated examples in Ceresa (1993), 34 had joined KA's while 18 did not!
Summary: Tyler's "KA" rule is good (with an exception or two) for the Scott 1-6 plain issue, but cannot be extended to the overprinted values.

I hope this clarifies a bit Tyler's Batum pages. His discussion is basically valid (as one would expect from a careful and respected author as himself), but the "KA" rule is too inconstant to be used for genuine overprinted copies.
Fortunately, the Forged overprints give themselves away with their characteristic signs, so the "KA" sign is not needed anyway.

As mentioned earlier, I will blog publish a much more complete discussion about these fascinating forgeries in November.