Al-Tahdhīb likewise records:
“Al-Dāraqutnī said: ‘It is neither sound nor established.’ Al-Shāfiʿī said in the Sunan of Ḥarmla: ‘A person praying is not to draw a line in front of him unless that rests on a proven ḥadīth, in which case it may be followed.’

I say: Mālik stated in al-Mudawwanah, ‘Drawing a line is invalid.’ Among the later scholars who declared the report weak are Ibn al-Ṣalāḥ, al-Nawawī, al-ʿIrāqī, and others—and they are correct, for two defects prevent us from grading it ḥasan, let alone authentic: inconsistency (iḍtirāb) and anonymity (jahālah). Removing the first, as al-Ḥāfiẓ attempted in Bulūgh al-Marām, does not by itself remove the second, as is obvious; it seems he overlooked this when he deemed the report ḥasan, although in al-Taqrīb he himself acknowledged the obscurity of its two transmitters, Abū ʿAmr ibn Muḥammad ibn Ḥurayth and his grandfather Ḥurayth. None is protected from error except the one whom Allah protects.

I have explained the two defects in detail and cited the scholars who weakened the report in Ḍaʿīf Sunan Abī Dāwūd, no. 107, and earlier I noted that Ibn al-Ṣalāḥ used it as an example of a shādh (anomalous) ḥadīth in his introduction; see Principle One.

Al-Bayhaqī’s phrasing quoted by the author contains a subtle indication of weakness, for he limited his comment—“There is no problem with it”—to “for this ruling,” as though he regarded the report as belonging to the category of virtuous deeds (faḍāʾil al-aʿmāl), where a weak ḥadīth may be acted upon. This is apparently the basis of al-Nawawī’s statement in al-Majmūʿ: “The preferred view is that drawing the line is recommended, because even if the ḥadīth is not established, it still secures a protective boundary for the worshiper. We have already mentioned the scholars’ consensus on acting upon weak ḥadīths in matters of virtue, not in questions of lawful and unlawful, and this falls under the former.”