That hasn't been changed.

Paul specifically said in a post when he changed ML that it now affects rarity and quality

So it shouldn't affect quantity of drops still.
So his data on amount of drops is either statistical error or a bug with ML

i dont think its a statistical error, i think thats just normal variance. everyone has seen how sometimes you go a full mana bar and dont get a single shrine.

normal. does that count under your statstical error?

i suppose it does, because he changed variables. being EE armors.

if he wanted to do this Perfectly, he should

1. Set of t3 magical gear with +3 ML

2. set of WHITE t3 gear, same types,

same set quantity of whistled killed monsters ( only way to accuractly determine the quality of the monster.

that controls for the highest amount of variables i think.

i didnt test amount of drops. i just stopped testing in one variation earlier than other one, as i got bored.
the normal vs magic+ split is statistically significant. also rare is significant. the mystical+ is not.

i didnt do perfect whistling indeed, as there is some noise (few EPs and some walking - normal monsters)

but the drop difference is clear. because i had 1-2 EP on both sides, not 4 on one and 0 on other.
the main thing that is different is weapon damage i had - few monsters did not die one hit kill - but not gonna repeat experiment now, maybe i do this later.

Really old post from Njaguar
https://forums.d2jsp.org/topic.php?t=3770403&f=272