Take for example "October the third, two thousand nine" (10/3/2009, MDY). In German one would pronounce this as "der dritte Oktober zweitausendneun" (3.10.2009, DMY). In Hungarian "kétezer-kilenc október harmadika" (2009. 10. 03., YMD). All of these pronunciations reflect the order numerical dates are written: using any other convention would simply make them less natural to pronounce.
Now it is certainly not unheard of that the order of pronunciation also changes (see the British "the third of October"), but changing language is a lot harder in some cases impossible without breaking grammar than changing written numerical representations.
Sure, date order may be decided based on accidental grammatical features of the target language. But that just reaffirms that "practicality for the common man" is rationalization.
Naturally, the written convention need not affect pronunciation: in the case of numbers, the world standardized on decimals with place values increasing from right to left, but most languages retain their own idiosyncratic ways of naming/reading them This causes negligible friction in practice.
Why big endianness is great though, in practice is that most of the time you actually do care about the year when you use the written form AND it's unambigous. The problem with the reverse notations is not that it seems illogical to some people. The problem is that both d/m/y and m/d/y is used and you can't have the faintest idea which one you are looking at.
Now in theory we could have y/d/m, but we fortunately don't, so seeing the first number being a 4 digit one, you know how to interpret the date.
I should note that the Hungarians (can't speak for the Chinese) obviously leave out the year when it's clear which year is meant.
The Hungarians (and presumably the Chinese) have their own just-so stories about why YMD is the most practical order: you use dates mostly when you sort heating bills, invoices, or even fruit preserves, and in all those cases you need to sort by year first.
Australians will tell you a tale about how "day before month" is human-scale, while "month before day" is not. Heck, ask a German why the decimal comma is superior to the decimal point, and you'll hear plenty of practical-sounding reasons. French carpenters will tell you how they use decimal units because those are much more practical than inches. British people will explain how left-hand traffic is the only practical choice (6a41ef).
People are much better at confabulating reasons for sticking with conventions than they are at telling practical and impractical conventions apart. Indeed, if there was one obviously practical choice, there would be no differing conventions for ISO standards to reconcile.