Well that may have been your experience. Mine is that processes are blocked while waiting for swap, so other process can run.

Plus even if a CPU intensive task is running it will only take 100% of the CPU during the time the kernel allows it. That's preemtive multitasking, as opposed to collaborative (cooperative?) multitasking as in Windows non-NT.

Of course the computer will be snappier with a dual-core than with a single-core. But generally one could say that single-core 2Ghz >= dual-core 1Ghz, because the single-core is used more efficiently (think of "turbo boost" in dual-cores).

As far as I can tell the only reason why we now have (N>1)-cores is because of power/thermal constraints. Following the example above, one can generally say that single-core 2Ghz uses more power than dual-core 1Ghz.

Well Linux is not real-time (at least by default) so this cannot be guaranteed (Symbian is actually real-time so it's not so affected by this).

In any case, the fact that the phone app takes a while to respond when receiving a call could most likely be solved (or alleviated) by tweaking the priorities and schedulers.

One thing that may also be worth thinking about is why many people expect instant response from their mobile computers (e.g. opening the calculator takes 1 second, etc.). I have a dual-core laptop (not state of the art, but "fast" for all I care) and things *do* take a while to start, even opening an x-term or a text editor. Why would I expect a resource-constrained device such as my N900 to be quicker than a full-blown PC?

Once you accept that there are limits to hardware *and* software, you may start accepting how things work, and actually start liking it as it is.