No, I get it. It has to do with bearing surfaces.
Tuning post as a clock face.
Imagine the bridge below us, below 6 o'clock. String is winding anticlockwise. The point where the string will sharply bend and insert into the tuning post, let's call it the break point.
Start with the tuning hole aligned vertically (i.e. 12/6) and pull the string straight through the hole. The string is not yet bent so no stress at the break point.
We start winding the string. The break point moves anticlockwise from 6 o'clock up towards 3 o'clock and as we do the string starts to bend at that point. String tension increases, with a force pulling downwards towards the bridge (parallel to 6 o'clock). The stress at the bend in the string at the break point increases too.
Once the break point of the string is at 3 o'clock, the tuning post starts to become a bearing surface that distributes the tension and takes some of the increasing stress off the break point. e.g. when the break point is at, say, 2 o'clock, the surface of the tuning post between 2 and 3 is supporting the string and therefore taking some of the force. The direction of the force is always towards 6 o'clock, so the bearing pressure is not evenly distributed across all points of the bearing surface but depends on the angle relative to vertical, i.e. bearing pressure increases from minimum (zero) at 3 o'clock to maximum at 12 o'clock and minimum at 9 o'clock (the magical 3/4 of a turn).
See?
