This damn EVA z

[ololololo]

Are those two distinct "letters", too?

No, because they correspond to what k looks like — it’s a capital letter. By the same logic, the letters h and b are the same; it’s just that h is an incorrect spelling of b.

Besides, a case can be made that those "mini-k"s are indeed just k gallows.

In the other two cases, this letter is in a position that is uncharacteristic for k — zepchy and tazain.
The position is uncharacteristic because a gallows is already standing nearby, and two gallows in one word are very rare.

Well, if it is a Latin month name, then the cipher can encode ordinary letter sequences with unusual sequences of Voynichese.

Generally, such a rare symbol, in my opinion, is not typical for Voynichese. This may suggest an homophonic substitution, where z is a variant for another letter.