You didn't understand the article I think. Yes, he says that larger calibers are more effective than smaller ones, but NOT that the deflection effect doesn't exist with bigger shells.
In other words, there's an actual effect INCREASING the armor effectiveness with sharp slopes whatever the caliber (a positive one, the OPPOSITE of your normalization theory), BUT this effect decreases with shell size.
First I don't agree with that article, normalization is not an effect imagined by WoT devs, it's broadly accepted and proved by actual test(like the 120mm penetration pic I posted).
Second, the article itself had been contradicting in using the term normalization:
“normalization” gets higher values with a greater shell diameter (that means more mass per mm²). In other words: Small calibers are worse against sloped armor.
If "nomalization higher" here means increasing the target armor, then the first sentence means larger calibre and more mass per mm² helps increasing armor? Even if so, why the second sentence says small calibres are worse?
Third, we are talking about obivious case that shell should penetrate despite angle, like tank AP hitting an APC. The force per space a tank shell apply to 10mm of APC armor is way beyond its resistence, even it's shot from 10 degree.
Things are not just simple trigonometric, if it is, did you ever see a paper bounce off bullet?