Well that's an even stranger argument. No developer worth the name would ignore a bug in their software because there's an easy way to close apps in the OS.
I wasn't using the term "bug"
only in the strictest sense of the word, but also to describe design flaws, like deep back stacks, or the fact that some apps just don't work as expected after switching back to them via the task switcher (like backwards navigation in IE, or not being able to return to an apps main page via the hardware back button). Developers ignore such "bugs" all the time, because Microsoft doesn't make it easy to find better solutions. That's not a strange argument, but common practice. Up until now, all the reasons mentioned for wanting a close-feature were design-flaw related, but for all those scenarios better solutions exist.
In this thread, I have for the first time heard somebody say they wanted a close-feature to dispose of WhatsApp, due to it crashing and becoming unresponsive. I've seen apps crash, but I've never heard of WP apps crashing in a way that left them hanging. A search on Google suggests that such behaviour is, if true at all, at least extremely rare. If bugs of that sort turned out to be far more common than I currently suspect they are, then I would change my mind and adopt your position (and adjust my opinion in regard to the potential of WP).
Any API that prevented screw ups to the extent you are proposing would be so restrictive as to severely limit the functionality of the apps that can be produced. And the WP API is already quite (IMO too) restrictive in what it lets developers do.
Are you seriously suggesting that no improvements can be made to how WP handles "fast resume"? Or deals with app state changes? Or that Microsoft couldn't deliver a UI component or two which offer a standardized way of navigation between app pages without piling onto the backstack? I see a lot of ways their APIs could be improved without limiting flexibility, and would also suggest that limiting flexibility in a few select places would actually be a good thing, like the depth of an app's backstack.
Perhaps your expectations are far too high. A smartphone is a handheld computer. It's virtually impossible to eliminate all bugs from any reasonably powerful computer system, or the programs that run on it. So having an easy way to close a misbehaving program (rather than a tolerable but clunky workaround) is actually good user-centric design practice.
Perhaps you are familiar with the difference between a computing appliance and a computer. I see a smartphone as the former. So does Apple. So does Microsoft. The idea is to limit some flexibility in exchange for reliability and ease of use. Seriously, if smartphones become as unreliable as an average consumer PC and require the same amount of administration to keep them running smoothly, then I'll stop using them. Adding said close-feature to deal with serious stability issues which an appliance should never have, is the opposite of good user-centric design practice. You're putting bugs at the centre of your thinking, not users...
WP and iOS are appliance OS', and they can achieve that level of stability. WP7 had already achieved it.