>The correct analogy is, “Suppose you were abused by your parent; should you be allowed to establish a benefit specifically and only for the abused children of other parents?”
That analogy is invalid because the original injustice here was discrimination, and people are proposing more discrimination in order to correct the original discrimination. Maybe that would be reasonable if you could be sure that the new discrimination narrowly targets people who unjustly benefited from the old discrimination. However, in practice this is unlikely to be the case: You'll have a situation where senior engineers benefited from discrimination, and we discriminate against a different set of junior engineers in order to "balance the scales". Two wrongs don't make a right.
Furthermore, as a method for achieving justice this is highly dysfunctional. There's no way to get consensus on what the "sentence" should be. There's no way to measure the degree to which the "sentence" has been meted out. It's just a big case of "squeaky wheel gets the grease". The more DEI professionals you hire, the more they will advocate for the need to hire DEI professionals, until the thing collapses into self-parody and Trump gets re-elected.
It's already possible to sue corporations for discrimination and violation of civil rights law. Why is this remedy insufficient? Maybe because there isn't actually a good legal case to be made that the alleged discrimination actually occurred, and people are just grasping at straws?
In any case: We can play these sort of zero-sum and negative-sum games until the cows come home. Functional societies don't cry over spilled milk, and instead focus on positive-sum games. To facilitate positive-sum games, we need a stable and predictable legal framework, not quixotic justice quests which mysteriously get ever more urgent the more the injustice recedes into the past.