Wow, this has struck a nerve.
However, as someone with a security background (sold my previous internet security company to IBM), I can assure doubters that Symbian has a much more sophisticated approach to security than Apple.
Even from an external perspective it's obvious that Apple is desperately trying to keep things simple for themselves (and thus, hopefully for the customer, although that doesn't always follow). The restrictions on third party apps are there for Apple's benefit (so they don't have to do the hard work Symbian has done and/or make their dev. environment as complex as Symbian's). The AppStore restrictions (eg. Google Voice being banned because of alleged pressure from AT&T) are similar, with Apple maintaining a single store worldwide so everyone is limited to the lowest common denominator.
This is an Apple tradition.
A Symbian tradition is exposing the complexity to the developer. 😉
The poster (Serious 60) who talks about a phone crashing when using an API without a required capability certainly points out the complexity of implementing fine grained capabilities, but it's notable that the capability wasn't allowed through, which would have been a security failure.
In other words, the poster hadn't discovered a security flaw, but rather a device flaw. (The poster even says so themselves, so I'm not quite sure what their point is.) Perhaps they're implying that this flaw could be used as an attack, but where's the motive apart form simple maliciousness? Not a good motive for the effort required to achieve widespread delivery. Note that the motive of the vast majority of malware is the same as the motive for most crime: filthy lucre (money).
So it doesn't make sense to follow this up with the rather naive comment about submitting a change to a core build perl script to insert a backdoor into the Symbian codebase. This shows a lamentable understanding of (a) how software engineering in general works and (b) how the Foundation's practices work (you can read about them -- in this case the package owner is responsible for reviewing all changes and it's incredibly unlikely that such a change would get by even the most distracted package owner).
Also, I have it on good authority (from one of our engineers who used to work in Symbian) that Symbian's processes were so heavy that they almost stifle innovation -- they are constantly striving to balance agile with robust (you can see that in David Wood's blog). I suspect that Nokia is a bit different, though...
In any case, we can all see what the Foundation's processes are, and they are certainly designed to ensure all submitted code is reviewed. (That doesn't guarantee anything, of course.)
Also a couple of other comments are amusing: the idea that hackers are not interested in Symbian is ludicrous. There are plenty of crackers out there working to pirate every third part app, and the installed base of Symbian is vast, so it's a very juicy target. It's just a harder target than something like Windows or even Mac OS. Things like the despised descriptors actually have a huge security payoff.
Finally, the comparison between the Apple AppStore and Symbian's failure (and let's be honest, it is a failure up to now) has nothing to do with security implementations, and Symbian's sophisticated security framework does the opposite of making it fun to develop for (which explains why there are not a surfeit of trivial apps -- it's just too hard to make an app of any kind).
Just some thoughts.