I think this contention in particular is a bit strange, at least to my mind - is your argument that Steiner should have mentioned it in his diary more than once? Or that it's unrealistic that they'd only reach the "noticeability" threshold once?And how none of these were seen by Steiner during the highly visible and extensive activity of Jackson's troops reported by him during the preceding four days leaves "more questions than answers".
Of course if they were part of the logs then they'd have only really passed through the town when the army marched.
(Personally I think Steiner's very reason for mentioning it was the contrast between Confederate objection to black soldiers and seeing black men as part of the AoNV; he may even have seen men carrying weapons, though one doubts they were formally enlisted as such.)
Something I want to point out though - since this came orginally from how McClellan's intel org should have known Lee had only a certain number of men - is that there is no single observation or combination of observations that can result in the picture of Lee having less than 60,000 men*. On the other hand, if combined correctly the historical observations can result in up to 104,000 infantry** plus however much cavalry Stuart had - and one of those historical observations is Steiner's 8,000 for DH Hill, which is the only observed value for that division and which misses out a brigade.
* except for the Baltimore American, which argued for there being only 5,000 - mostly cavalry - in the entire invasion force
** "Jackson and Lee" at Boonsboro on the 10th, "40,000 to 60,000" take that as 60,000
McLaws and Anderson, "30,000 men at Burkittsville"
Walker, "6,000 men"
DH Hill, 8,000 (Steiner)
So the low bar is 61,000 infantry, and the high bar is 104,000. Average the two and it's about 82,000, plus however many Stuart has.