Allo, I have to confess that I didn't read the article by Sheldon Brown. It looked lengthy and immoderately technical.
I still say that if you require more movement of the lever for any given movement of the brakes (ie -- the thingy dialled in) then you have an increase in modulation. Try setting your mouse to maximum sensitivity then tell me whether you think you have more modulation or less. When the pointer zips all round the screen at your slightest touch you'll have to agree that you have less modulation. Set the mouse for least sensitivity and you'll have to agree you have the greatest modulation. You can put that sucker anywhere (should you have the patience) but you have to move the mouse a great deal. You have sacrificed control for speed.
This is the same as you would do with the speed dial adjustment. The only difference is that a mouse requires the same force to move, regardless of the settings within Windows. With speed dial controlled brake levers you have a choice of moving the lever further, for any given braking force, increasing modulation, or you can move the lever a shorter distance, decreasing modulation.
You point out that screwing the speed dial in increases your mechanical advantage, and that is true, but user won't notice this. He might, at the end of a long downhill, notice less arm-pump, but the immediate effect of this change will be that the brakes feel softer or more spongy than in the top-most setting. I think you're halfway there, wrong only in believing modulation does not grow with mechanical advantage.
James.