It seems to me that it is a sort of blunder from Wittgenstein.
Wittgenstein criticizes the logical rules for identity already in 5.434, becuase they are not expressed with a "correct logical notation".
It seems to be a critique of Frege's and Russell's theory of quantification and identity. Wittgenstein’s approach seems to be that no adequate logical
notation would include the identity sign, and to claim that the "=" sign becomes unnecessary if stay consistent with the use of names:
5.53 Identity of object I express by identity of sign, and not by using a sign for identity. Difference of objects I express by difference of signs. [see also 2.0233]
Thus - it seems - a logical perfect language will uses different signs for different objects, like the numeral 1 to denote the number 1 and the numeral 2 to denote the number 2. If so, a formula like 1 ≠ 2 is a non-sense, or at least "useless".
According to Wittgenstein, the only legitimate use of the sign of identity
is at a meta-level, in order to talk about the use of signs, and not to assert anything substantive about the world. Thus he says:
4.241 When I use two signs with one and the same meaning, I express this by putting the sign ‘=’. between them.
So ‘a = b’ means that the sign ‘a’ can be substituted for the sign ‘b’.
(If I use an equation to introduce a new sign ‘b’ laying down that it shall serve as a substitute for a sign ‘a’ that is akeady known, then, like Russell, I write the equation - definition - in the form ‘a = b Def.’ A definition is a rule dealing with signs.)
4.242 Expressions of the form ‘a = b’ are, therefore, mere representational devices. They state nothing about the meaning of the signs ‘a’ and ‘b’.
This account of the role of the identity sign is in contrast to
that of the mature Frege, who initially adopted something like Wittgenstein’s view himself in his Begriffsschrift (§8) :
Identity of content differs from conditionality and negation in that it applies
to names and not to contents.
But later Frege rejected it in Über Sinn und Bedeutung. This rejection were motivated by the fact that the metalinguistic account made identity statements in general (and mathematical equations in particular) into relatively trivial linguistic truths, whereas really they were capable of expressing "real"
We can consider the logical axiom for identity:
according to the standard semantics for first-order language, it express the "trivial" fact that "every objcet is equal to itself".
By way of the quantification axiom : ∀x α → α[t/x], where t is a term, we can derive e.g. its "arithmetical" instance : 1 = 1. Again a "trivial" true sentence of arithmetic : "the number 1 is equal to itself".
But the properties of identity are used also in the arithmetical axioms [see Peano axioms] for the successor function S and for the sum (+) and product (binary) functions.
With them, and the usual abbreviation for the numerals : 1 for S(0) and 2 for S(1), i.e. S(S(0)), we can derive the formula :
1 + 1 = 2.
This formula can be "read" at the meta-level (as Wittgenstein do) as expressing the identity of reference between two terms (two names).
But it express also an arithmetical fact (as Frege stressed) that is not a "linguistic" one, but a "real" piece of arithmetical knowledge.