People with these statistics will be making things better for themselves and for advertisers. This may often coincide with making things better for users, but the type of data you're talking about is mostly useful for making it harder to skip ads.

I like my free, RSS-shared MP3 files as much as the next listener, but these things aren't free to produce, just like PocketCasts isn't free to make. Podcasters sell ads, and advertisers rightly want something in return. If ads are easy to skip, they have no assurance that they're getting anything for their money, so the money dries up, and the podcast goes under. There is always going to be a game of back and forth between ads and avoiding them, and this is just the next move, if it even happens.

I prefer the direction it's going in now where most podcasts I listen to have an option to subscribe to pay the podcaster more directly. This does not require turning my player into spyware yet it keeps the podcasts going.

I agree in theory but what else can be made better about podcasts? Is it worth the trade off for everyone's data? These are questions I believe were asked and answered with NPR One. Also, anonymized data is still data and that derived information can be very powerful.