this hadith is weak and misinterpreted. the concept of 'majhul' (unknown) applies here. i will try to elaborate a little more, insha Allah.
though Darussallam has graded this hadith (sunan ibn majah 1986) hasan, inclusion of Abd al-Rahmān Muslī in the chain of narrators makes this hadith weak, if not very weak according to many scholars. hadith master al-Maqrizī in his Mukhtasar Kitab al-Witr (p.50) mentions this.
Abd al-Rahmān Muslī who not only is 'not known' as stated by al-Dhahabī in al-Mughnī fīl-Du‘afā’, but is not known to narrate anything else whatsoever in all of hadith literature!
thus as per the criterion by ibn hajar in Taqrīb al-Tahdhīb one from whom only one person narrated, without declaration of his being trustworthy, is majhūl (of complete unknown reliability).
- [ i do not find it relevant to mention in detail here about the mistake ibn hajar made in grading Abd al-Rahmān Muslī 'magbul'. please refer to Tahrīr Taqrīb al-Tahdhīb by al-Arna’ūt and Ma‘rūf for further clarity if you need any ]
furthuremore this perticular narration is recorded in ibn majah, abu dawud and ahmad all of them through Abd al-Rahmān Muslī.
as you can see below in the grading of Darussalllam, they have graded the hadith in musnad ahmad 'daif' because of the same 'majhul' narrator but not followed with the same pattern in the narration in ibn majah.
now to the narration in sunan abu dawud and sheikh Albani has graded it 'daif'.
finally, the sense of the words 'a man is not to be asked why he beat his wife' does not mean that he is considered free to do so, because he is not, since the Prophet (S.A.W) explicitly prohibited wife-beating but rather that one should not be pressed to disclose anything blameful about one's wife.
in conclusion, we should not try to derive any ruling from weak narrations unless there is enough corroborating evidence to support it and that is not the case here.
may Allah the mighty and sublime grant you mercy and guide you to the straight path.
Allah knows best.