The better dies are likely more consistent, with fewer 'lemons.' Like you say, better on average. But it is surprising how well the cheap die occasionally performs. You just never know till you check.
Also, the experiments really showed me how many different parts of the loading process affect the final product. One flaw in any of them can screw up the end product, yet without the right tools you don't know where the breakdown occurred.
Personally, I'd rather discover problems at the loading bench than after I went to the range. Far more cost effective.