It would make sense that anything other than normal mode was not really playtested in the strictest sense. They are likely to have run all skills just to check if they are working at all, but balance-wise, they would have concentrated on normal mode (despite claiming it to be "tutorial mode") because they assumed the majority of the playerbase would be playing that until they were ready for the first batch of balance changes.
Those few sad people (compared to the bulk) that have crossed the threshold to the nightmare lands and the hells beyond are nothing but balance-beta testers at this point.
It only turns out that normal mode is so god-awfully easy that even the bulk of the playerbase that is largely comprised of what one could not describe as "canny gamer" are rushing the content, so those balance changes better roll in fast.
How do I arrive at that conclusion?
Mainly, scaling issues.
Mana works in normal. It still works to a degree in nightmare, stops working properly in hell. (Really, have you ever compared mana to arcane power in normal? No real difference there)
Pets work in normal. They still work to a degree in nightmare. They don't work in hell (unless providing background color with their bodily fluids is "working as intended")