"Commentary manuscripts contain, apart from the commentaries, the continuous NT text. It is unlikely that these commentary manuscripts served as exemplars for manuscripts containing the continuous text. But it is very plausible that that text form was chosen as the basis for a commentary which was the most highly esteemed at that particular time and place. Accordingly, [that text-type] would also have been used as an exemplar in the scriptoria."
That is, the text chosen for the commentary would be the one popular for regular copies in that scriptorium where the commentary was made, at the time it was made. It would not normally be an "ancient text", but rather an 'approved text'.
This is an important observation regarding the conventions for such commentaries, because the discussion in the commentary itself would often be incongruous with the text presented alongside it. This is because the commentary would often have been composed some 5 or 10 centuries earlier than the commentary manuscript, and the author's text would differ from the one popular many centuries later. The commentaries would not be rewritten to correct this, and in fact it would probably not be noticed, as the two tasks were done separately in sections, even when written by the same scribe.
To take a specific case, 'Codex X' is not an ancient Uncial manuscript, in spite of its conventional letter ('X'). It is actually a late 12th century commentary, with gospel text included. The commentary and text is alternately placed in sections in the manuscript body.
Only parts of text which will be relevant to the commentary will be copied alongside the commentary. Because of this, there is a perfectly obvious explanation as to why this manuscript does not contain the Pericope de Adultera (John 7:53-8:11). The ancient commentary doesn't discuss it, because this passage of John was not read publicly during Pentecost services. The commentary follows the sections of John as they are read in public worship, so the commentary doesn't comment on a passage which is left out of the services.
The scribe who copied the sections from a continuous manuscript into the commentary did not bother to copy Jn. 7:53-8:11, since it was not relevant to the commentary. So even though this commentary manuscript generally gives a faithful copy of the continuous manuscript used to make it, the commentary manuscript itself (Codex X) cannot tell us what the original continuous copy had or did not have in its own text.
It is wrong to call this commentary a "continuous text manuscript". It is not. It is a commentary which contains text from a continuous text manuscript. But it cannot tell us more about that lost exemplar than it has bothered to record. As it is, the passage up to John 7:52 is copied for one section, and the passage beginning at 8:12 is copied a page or two later for the next section of the commentary. We can never know from this what the original continous-text manuscript looked like at the critical junction, 7:52/8:12. It may or may not have contained John 7:53-8:11. If it did, the copyist of Codex X did not record this fact for us, as it was of no concern to him when constructing the commentary with text.
In view of this, it is pointless to cite "Codex X" as if it were a witness for or against the text of the Pericope de Adultera (John 7:53-8:11). It is not a 'continuous-text' manuscript, except in the sense that it contains text from a real 'continuous-text' manuscript. It is not even a NT manuscript at all; it is a commentary, which happens to contain text from all four gospels, but not entirely complete.
Much confusion regarding the 'witness' of Codex X would be eliminated if it were simply dropped from the apparatus, as it has nothing to say about the authenticity or position of the Pericope de Adultera.