How does VGA manage to show 720x400 resolution text on a 640x480 display?
This might be your basic misunderstanding here. Displays as used back then and what VGA was designed for, are not a fixed 640x480 device, but analogue CRT (Cathode-Ray Tube) monitors. Fixed size displays only became a common thing for desktop PC way later with the advent of fast, high resolution LCD screens (*1).
Number of pixel per line is thus only limited by output speed of the computer - and beam manipulation speed by the CRT. Likewise the number of lines. Due the way of the receiver of a CRT is build, it will adapt. This is called synchronisation (syncing). A 15% higher pixel rate is quite within reason for next to all CRT.
So in a way, yes, the pixels were a bit smaller. Except this was not reached by dropping parts of them or calculation some intermediate values like it is today, but by literally making them smaller in width. Which the CRT would show faithfully (*2).
Later, when Super VGA came, Multisync CRT appeared. They were able to sync to quite wide rates from 320x240 to 1600x1200 resolutions, some even beyond that.
*1 - Well, Laptops and other special purpose devices had them first, but they were not driving the standards back then.
*2 - There are limits due the size of the colour mask, but again they will not become an influence at variations that small.