I therefore say: the mursal (broken-chain) and mawqūf (Companion-stopped) versions do not, in my view, invalidate the first fully connected marfūʿ (traced back to the Prophet) report, for its transmitter relates additional information and, as the hadith principles stipulate, what he adds must be accepted. Moreover, the wording of the mursal narration itself shows that it is an independent report, entirely separate from the marfūʿ version, because it is extremely brief.

It appears that whenever Abū al-Jawzāʾ narrated the report as marfūʿ he transmitted it in full, mentioning the occasion of revelation; but when he cited the verse merely in explanation he related it in abridged form, without raising it to the Prophet or ascribing it to Ibn ʿAbbās—even though he had originally learned it from him. Hence the mursal transmission does not undermine the connected one. Allah knows best.

As for al-Ḥāfiẓ Ibn Kathīr’s statement in his Tafsīr, “This ḥadith contains severe irregularity,” I do not concede that point. The person who was looking from beneath another’s armpit could well have been a hypocrite or an ignorant Bedouin; this is obvious and removes any claim of anomaly or difficulty. For that reason, we find no one who produced or cited the ḥadith describing it as gravely irregular—not even al-Ḥāfiẓ al-Dhahabī, famed for his meticulous textual criticism. On the contrary, as you have seen, he declared it sound; and Ibn Kathīr himself says of him in his History, when mentioning the year of his death: “With him the leading shaykhs and masters of ḥadith reached their culmination.”

The ḥadith is also proof that women used to pray behind the Prophet ﷺ with their faces uncovered, and this is supported by the ḥadith of ʿĀʾishah (may Allah be pleased with her): “The believing women would attend…"