This is what I mean by "evidence". The reason I ask is that most developers (other than publicly held companies) don't discuss these details with the general public.
Kidding aside, one other aspect to consider is the amount of investment required to get an application released on either platform (the true profit vs. just revenue). A game like iShoot can be put together by a single developer using commodity hardware in less than 100 hours. This application, once released incurs no maintenance cost in order to generate revenue (see note below). A developer can therefore turn out additional applications of this class with little support overhead required.
I haven't written a Facebook app yet, and perhaps a successful Facebook application can be created using the meager developer resources described above. However in addition to initial development costs, as I understand it, the developer must provide the infrastructure to run the application (web hosting, bandwidth, etc.). On an application which receives the volume of users you describe, this can add up to a significant ongoing expense.
You clearly know more and have more experience with the Facebook platform than I do, if I am incorrect in my statements please enlighten me.
*of course you can continue to spend money on promotion and advertising, but since this applies at least equally between both platforms, I consider it implicit.
We didn't have to worry about being accepted into any store. (Facebook has an app directory that takes a couple days to get reviewed for, but honestly there's very little benefit to even being in it.) After we launched, upgrading was as simple as any website. We didn't have to make some fairly hard to achieve top 25 list to get traffic, in fact, we're hoping for the reverse.
Developing for Facebook isn't that different than any other web development (in fact a lot of apps run in an iframe). It's far easier, and gives far less control to a third party. It's far more lucrative, doesn't have the "hit or miss" aspect of the app store.
You're right about the hosting cost. Only some iPhone apps need that, many do not. All Facebook/Myspace ones do. But the greater monetization potential more than makes up for it if you're doing games (I'm not so sure about other types of apps). eCPMs on offer pages are measured in the $100s. Many people I know (in a way that would be admissible in court) are pulling in north of $500. Even if that only amounts to 1% of your traffic, that's still ludicrously high, especially on the volume you can get very quickly.
The only way in which it is inferior is it isn't sexy and nobody is going to write lots of articles about you. The iShoot guy got more coverage in bigger media outlets for bringing in $600k than Zynga did for making $40-$50m last year.
To put it in perspective, I'll probably bring in more rev than iShoot this year, and I'm nowhere near Mob Wars in traffic numbers. There have to be hundreds of guys out there bigger than me. And unlike iShoot, I'll keep doing it indefinitely, not just until I fall off of the top 25 (which I never even had to hit).
When it comes to third party apps, Facebook is really the leader right now. In fact, I'd argue that in the tech world, they're setting the terms more than Apple at the moment, just in a different sphere.