And you don't understand that this is actually a matter of playing with the definition of "qualified".
Except you are taking the word "qualified" and trying to make it an idol. You need electricity work done on your home? Let's grab 100 random people and find out what they do for a living. Looks like 3 are electricians. Tough to choose. I mean, we've got 97 people that don't know how to do electrical work and 3 that do. How are we going to select from this group of 100? How do you pick one?
You are wanting us to believe that we are just randomly grabbing people off the street and tossing them into positions.
I am not saying that at all.
The reality is that instead of choosing 100 random people, we are selecting 10 random electrical contractors. And then choosing from their qualifications, reviews, etc... Of course, picking "the right one" isn't exactly black and white.
When you have 10 it probably doesn't make sense to apply hard filters anymore. But what if you had 10,000? (By the time you're looking at a VP or SCOTUS pick that's not an unreasonable number.) Do you compare them? Or do you filter the pool by some means?
DEI is by stopping filtering at the point where filtering more would exclude all of those with your desired attribute(s), then claiming everyone still left is equally qualified. And, realistically, VP picks normally are. They aren't being selected as the most qualified, but as the most qualified that appeals to certain segments.