Maybe if they would look at the feedback they got on the ads (yep, people taking the effort to click on that tiny tiny x in the top right corner of the ad) where people said that the ad were irrelevant or even offensive. Maybe if they started with the assumption that these people knew what they were doing and hadn't managed to hit that specific combination by random chance and especially not three times or more in a row. And maybe if they had removed these users from the group that got said ads then that could be a good starting point for further work.
I'm no machine learning expert by any means but taking into account the current situation there should be some room for improvement.
If anyone missed it: I'm sarcastic. I believe ads targeting are either out of control, or Google as a company aren't as smart as they think they are or they have a really deep disregard for users.
Besides, Google etc. turned me off ads years ago, so they didn't work diligently enough.
Its not customized ads to YOU, but its customized ads to a particular audience, which may or may not include you. That audience might include a bunch of people who are extremely likely to buy things, and it definitely moves the scales for those companies.
Google and Facebook work diligently and at great expense to make sure that not too many people are "turned off" by ads - an example is even recently how Google rolled back its search results ad UX change after the outcry started to pick up stream. It would be an existential threat for them if everyone started moving towards ad blocking. So they simply keep people from getting to that threshold.