I don't see any mistake.
1. Word; A relative reference to the word you wish to edit
2. n: = LEN(Word); the number of characters contained the Word
3. aRow; a relative reference to the entire row containing the active cell
4. k: = COLUMN(INDEX(aRow,1) : INDEX(aRow,n))
Note; for me this is simply=SEQUENCE(1,n)
5. chr: = MID( word, k, 1 ); the characters as an array
6. selected: = IF( COUNTIFS(aux, chr ), "", chr ); array with unwanted characters extracted
7. = CONCAT(selected); Worksheet formula giving result
It depend on one's understanding of your requirements. I took it that the initial A of "Armazem" should be eliminated because "a" appears within an auxiliary cell. If it is alternate characters you are looking for, you have already provided an answer in your auxiliary table. My approach (modified from previous) that achieves your desired result has
1. Word; As before
2. n: = CEILING( LEN(word), 2 ) / 2; the number of characters to be returned from word
3. aRow; as before
4. k: = 2 * COLUMN(INDEX(aRow,1) : INDEX(aRow,n)) - 1; sequence of alternate numbers
Note; for me this is simply=SEQUENCE(1,n,1,2)
5. chr: = MID( word, k, 1 ); the characters as an array
6. = CONCAT(chr); Worksheet formula giving result