As a developer, you choose the platform you intend to use to deliver your application. Nobody is being forced anything. There is a huge potential market outside the App Store.
If you bet the company on Apple's good mood, it's only your fault they pull your app, change the rules of the game and drive you out of business. They own the clients. If you want to sell to their clients, you better be ready do jump when they tell you to. Actually, it's no different than selling to a telco on a platform like Brew.
As a dominant player, Apple not only has to foster the development of killer apps for their platform, they cannot afford killer apps evolving on other platforms. Unless they change course, that's what will happen.
Development for a mobile device is different. If you were developing a Windows application, you'd be expected to own a Windows system. If you're developing a MacOS X application, you'd be expected to own a system that runs MacOS X.
But what about an iPhone application? By the above logic, you should only be expected to own an iPhone. Now, this analogy breaks down a bit, because I wouldn't expect anyone to actually develop a serious piece of software directly on an iPhone (at least not without an external monitor, mouse, and keyboard). So you need another platform to do your development on. Apple is requiring you to buy a Mac. Clearly they have development tools that work on Windows, otherwise Win Safari wouldn't be possible. A port of the iPhone SDK to Windows (or Linux, or whatever) is certainly not out of the question... it's just a matter of throwing resources and time at it. Now, again, I can certainly believe that Apple would think that restricting iPhone development to Macs is in their best interest, but it's ludicrous to suggest that this decision is also in their developers' best interests. I think that's what a lot of people are pissed off about.
Go to another mobile platform? Sure, you can do that, but then you're restricting your market. Yes, there are quite a few other lucrative mobile platforms (several of which have greater market share than the iPhone), but, as a developer, you end up with the unfortunate choice of having to either not develop for the iPhone, or buy an expensive piece of hardware that you only use for iPhone development. And by "expensive" I don't just mean the price tag. I also mean costs in training (even if it's self-training), maintenance (in 5 years your speedy Mac won't be so speedy anymore), and upgrades (IIRC you already can't install the latest version of Xcode on anything older than Snow Leopard; I imagine this trend will continue).
I know people like you seem to think that it's an easy binary decision: 1) don't buy a Mac and don't develop for the iPhone, or 2) buy a Mac and develop for the iPhone, but there's a lot that goes into that decision, including a judgment of whether or not developing an app that you may not be "allowed" to distribute is worthwhile.
It's not so cut and dried.
As with any business decision, you have to weight as many factors as possible. Unless you have some marketing muscle for promoting your app, chances are it will drown in an ocean of similar applications. Worse: in an ocean of applications with similar names and descriptions.
Going the Apple way has never been an easy thing. At least in the pre Xcode days, Macs had a wider choice (like two vendors?) of development tools. I never liked MPW and preferred the Metrowerks tools. On the Mac, you program to a small slice of the pie, but a less crowded one. The iPhone is different - it represents a huge slice of the "people willing to pay for mobile apps" pie, but the slice is very, very crowded.
If you want a Mac, go buy one. You don't need Apple as an excuse. They are very good computers, very well built and have an OS that, IMHO, is second only to a decent Linux box as a development platform.
It's so good it breaks my heart to install Linux on top of it. So much in fact I never did it.
Except once, when I installed MkLinux on a Power 9500. Very nice X terminal, BTW.