This is the argument that gets made, but very rarely is it true.
From Neumann et al in 2019 [1]
"When investigating gender (being male) and age (three different tiers: 18-24, 25-34 and 35-44 years) individually, we find that digital audiences for gender are on average less often correct than random guessing (accuracy of 42.3%)."
If the accuracy of targeting is worse than random guessing on average then it's value is less $0.00. Advertisers would reach more of their target audience by simply buying more media instead of spending money on 'targeting' even after you discount wastage to $0.00 in 'value'.
I agree with everything you're saying about programmatic *in theory*, but I would argue that in practice the whole system is just broken.
[1] https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3203131