The letter is a good place to state plainly what the atlas is doing when it lists forty four Glagolitic letters. That list follows the arrangement of the Unicode Standard, which encodes the signs attested across the tradition, including regional and positional forms that no single manuscript uses all of. It is a repertoire, not a reconstruction of the alphabet as Constantine made it.
The oldest Slavonic account of the alphabet, the treatise on the letters by the monk known as Chernorizets Hrabar, gives a figure of thirty eight, and that figure does not match the number of distinct signs in the surviving codices, nor do the codices match each other. No manuscript survives from the mission itself. The earliest Glagolitic books are a century or more later than the alphabet, they were made in a different country from the one the mission served, and they differ among themselves in which signs they carry and in what order an abecedary lists them.
So a variant such as initial izhe sits in a grey area by construction. It is a real form, used by real scribes, and whether it counts as a letter depends on what is being counted. Naming it and saying why it is uncertain is more useful than deciding the question by leaving it out.