Money to pay the artists listened to by free users needs to come from somewhere. I don't think artists and rightsholders are willing to accept a lower per-play revenue from free users. Which means I doubt that the ad revenue is enough to support the free accounts without subsidy from paid users. (The ad revenue isn't nothing, and the ads helps encourage conversion to paid user, so they're not without value to Spotify, but I suspect actual paying customers are worth more than free users even though they stream more.) So if you take these two assumptions:
* A stream from a free user pays the same as a stream from a paid user, and
You could, I guess. Why would you want to, if you're Spotify? It's more bookkeeping for no benefit to them. The author believes he's getting hosed because gyms and nail salons are playing Spotify all day over their PAs, but... I'm not sure that's true, and even if they are, those are still plays Spotify has to pay for.
I'm almost certain free users pay out a lower rate, payout information from Zoe Keating has confirmed that there is a somewhat large disparity between rates even within the same country with no explanation attached, there's only 1 or 2 things it could be. Radio payouts are lower because it constitutes a 'radio play' and therefore gets less than a premium user playing the song on demand so there's definitely variance within the payout rates.
I think there certainly is a lower rate for radio -- I think Spotify gets Pandora-like rates for Spotify Radio, which I believe are lower than the rates they pay for their normal music streaming. I don't think there are different rates for free and paid streamers, though.
People make tools to game usage-based rewards in free to play games. They call it "idling". You could hack this by making an idling tool for spotify that would re-stream your tracks when you weren't using spotify.
* A stream from a free user pays the same as a stream from a paid user, and
* Paid users bring in more than free users
I don't think Vulf's idea works.