I suspect disguising oneself as a man worked better when fashions were absolute. I mean, if the trousers or bonnet or whatever were as clear a gender marker as a beard or breasts, why would you question it?
Sure, these days we see both genders in all styles of clothing, so we're used to more than one look. But it would be different if you had flat assumptions:
1. All soldiers are men.
2. X is a soldier (complete with men's uniform, etc.)
3. Therefore, in the absence of obvious markers like breasts, X is a man.
Or ditto a beardless person in gown, apron, bonnet, et al; of COURSE she's a woman, you maroon!
Even when you have cross-gender disguise plots in books earlier this century, they tended to make flat assumptions about hair length. "The boy's cap was knocked off, and the sudden spill of long hair revealed to us that he was, in truth, a woman!" This gets particularly silly when you're talking about non-Western cultures; you'd think even an Englishman could imagine long-haired men from Tibet or mystical barbaric forests. But no; pre-flapper, OF COURSE all women had long hair, unless they 1) had been sick, or 2) had been shorn in disgrace.
But we can theorize all we want. The truth is, there are several historical cases where women passed as men for decades, even among those who knew them well. So it must have worked somehow!
--Nonie