I think that means we're doing it wrong, but nobody wants to admit it.
For example, I buy hot rod magazines for ads for cool stuff for my car. I'm not interested in furniture ads in those magazines, and furniture makers don't advertise in car mags. When I want to buy furniture, then I'll buy a home decor magazine and look at furniture ads. I don't want to see a fuel injection pump in a home decor magazine.
Hemmings Motor News is another example of effective advertising. It's nothing but car ads - and people pay for a subscription.
An online example is newegg.com. All it is is computer electronics. It's my go-to location when I'm looking for stuff for my computer. Seeing ads on newegg is exactly why I'm going there.
Of course, if I actually was able to buttonhole some guru at Google who understood the business and my situation, they would just look at me like I was an idiot and say "we warned you that if you turned off the various personalization features the ads wouldn't be relevant".
But if I left every option turned on, the ads wouldn't be better, just creepier.
People like reading ads about products they're interested in. I know I do. This idea that ads are all toxic because their purpose is to sell just isn't right.
I think most people back in the day came away from these magazines wanting more things they didn't even know existed prior to opening it up.
Back in the 1980s (in the Datalight/Zortech days), the only game in town was advertising in computer magazines. We knew it worked because the day the monthly mag appeared on the shelf, we'd get a big spike in sales, that would then taper off through the month. It was fairly easy to correlate which computer mags mattered, how ad placement in the mag affected results, and how the size of the ad worked out.
For example, we learned to never bother advertising in August. Programmers apparently all went on vacation in August :-) but September would be a huge month for us.
I tried print advertising in the aughts, but there was literally zero return on investment.