In short, citing this report as proof that the knee is part of the ʿawrah is untenable for three reasons:

First: It is mawqūf, stopping with the Companion.

Second: It is set against other reports that are more authentic.

Third: It was spoken about the ʾamah (female slave); hence it is more restricted than the general ruling. Scholars have in fact differed greatly over the ʿawrah of a female slave, and the view whose evidence is strongest is that she is no different from a free woman.

Al-Ṭaḥāwī’s procedure—analogising the free man and others to the female slave so as to declare the knee ʿawrah—perfectly mirrors al-Nawawī’s analogy that a man’s prayer is invalid, like a woman’s, when any part of the ʿawrah is uncovered. We have already refuted that at sufficient length.

It may be for the very objections just mentioned that the author of al-Hidāyah and other jurists ignored the ḥadith of Abū Mūsā and instead relied on the report, “The knee is ʿawrah.” Had that report been sound, it would indeed have furnished them with clear proof; but it is not sound—on the contrary, its weakness is unanimously acknowledged. Al-Dāraquṭnī transmitted it through the chain of al-Naḍr ibn Manṣūr al-Fazārī: “I”—meaning Abū al-Janūb, whose name is ʿUqbah ibn ʿAlqamah—“heard ʿAlī (may Allah be pleased with him) say…,” and he related it as marfūʿ.

Al-Zaylaʿī states in Naṣb al-Rāyah:

“Our shaykh al-Dhahabī says in his Mīzān: al-Naḍr ibn Manṣūr is feeble; Ibn Ḥibbān said, ‘He is not to be relied upon.’ As for ʿUqbah ibn ʿAlqamah, al-Dāraquṭnī and Abū Ḥātim al-Rāzī both graded him weak. The compiler repeated the report in the section on Disapproved Matters, attributing it to Abū Hurayrah, but we could not trace it from him. In al-Imām, Abū Ḥātim al-Rāzī said: ‘ʿUqbah’s narrations are weak, and al-Naḍr ibn Manṣūr is unknown.’”