According to this, Facebook revenue per-user-per-year is about $30, so $2/mo (post-Apple cut) is probably not far off for Twitter: https://www.statista.com/statistics/234056/facebooks-average.... But of course, why replace that revenue when you could double it?
That $30/year figure is likely averaged over the globe, therefore their North American stat is likely quite a bit higher. Given the quality of ads on twitter that I see, I suspect my value to them is much, much lower. Their ad network just seems terrible compared to Facebook so it's surprising they didn't offer a $10/month ad-free version.
you should really get these kind of data straight from the source when you can. stastica is sometimes useful for some hard to get metrics. this is not one.
edit: here's an ARPU estimate for Twitter in 2016, it was around $2/quarter: https://www.statista.com/statistics/430874/twitter-annualize...