Hmmm.

Well as far as the Z600 goes, it will not be useful in a smartphone just because it has lower battery consumption (the difference is anybody's guess) than your typical 1GhzA8 (eg Hummingbird) and the performance gained is only slightly higher than the Hummingbird (maybe the same if hummingbird is overclocked to 1.5GHz).

And what strengths does having the Z600 over the Hummingbird give you? Actually nothing, since Z600 is not x86 compatible (at least software-wise). Actually it disadvantages you since the Z600 is still larger, so either the device will be larger or the battery will be smaller compared to smartphones using A8 or A9 processors.

But on the tablet front it is a completely different scenario.
Why? Well the portability (or size) of both platforms in the larger shell is very close that it no longer matters. Next important thing is that the battery is much larger so the normal/idle battery life difference is again too small to consider, it now depends on which company can optimize the software best.
But where there is a difference is in performance. As stated, an overclocked 1.5GHz (single core) Cortex A8 with a dedicated PowerVr GPU should be equivalent to the performance you can squeeze out of the Z600. But what does change is when you have Cortex A9 cores (these have the same battery draw at the same GHz as the A8's except they have a more performance = effecient). So lower-end A9's (1.2GHz) would be able to match the Z600 fairly easily, but when you have them pushed hard (2GHz) you are surpassing the Z600 and when you double-that to a dual-core ... you're leaps and bounds ahead of the Z600.

What no-one knows right now is the battery life difference between the Z600 and a 2GHz-dualcore-A9?

But if Intel is to be successful in the tablet space it must either:
1) match the performance and battery life of said processors which is a difficult task (due to x86 constraints)
2)OR it must somehow convince Microsoft to port Windows (7 or its successor) to the Z600 (or its successor) and this is PROBABLY not going to happen (in fact you can forget about it)
3)OR Intel must be able to design a new Atom/core-i2(?) line which can provide more performance than the N450 (like an SU7300) at the same battery life of the N450 BUT have it in a small package where it can fit with a 6-cell battery into a tablet about the size of the iPad
4)OR A wealthy company must accomplish a tablet with the features of "bulletpoint 3" and it have a completely new overhauled skin over Win7 (so that it is finger and tablet friendly) and it must also become popular and attract other OEMs to do so ... so far only the ExoPC is heading in this direction but it isn't doing a terrific job of it. BTW this is the easiet way Intel could compete in the tablet industry but your regular OEMs (HP, Acer etc) are not interested/not offering one any time soon.

Didn't Intel sell those licenses to Marvell?

Yeah, I'm very happy with my AMD 6core with an overclocked GTX 460 for $1300. Intel would have added a few hundred to that.

You have the compare the same targets - Intel is not sleeping, the real competitor to the A9 is the Medfield, and that's dual-core 32nm stuff that actually tries to be a SoC, so a completely different story to Moorestown (which really is for tablets at best). Also, comparing the lowliest Z600 to a super-clocked A8 is not really meaningful (that 1.5GHz A8 under load might even suck more juice than the 1.2GHz Z600). The devil is in the details, so I don't know whether Medfield will be up to the hype, but the question is still the same collision course one - what will happen sooner - ARMs reaching (THEN equivalent) Atom performance or Intel chips reaching (THEN equivalent) power high-end ARM consumption (even though the itch is to shout ARM ! the answer is actually not as clear cut and will actually depend on software a great deal).

Yes, Linux. But not necessarily Maemo or MeeGo. Anything as long as it isn't Microsoft Windows.

IOW, Android, ChromeOS, iOS, ...

It's worth keeping in mind that Atom is still at 45nm even though Core is at 32nm. In 2011, that'll become 32 and 22. If/when mobile devices become critical for Intel, they'll be able to do much better than they are now, relatively speaking.

When the going gets tough, Intel will buy their competitor(s) to extinguish their superior technology and continue use its own bloated x86 architecture.

Considering all the anti-trust scrutiny Intel has been under as of late, I don't think they're going to make ridiculously anti-competitive moves like that.

Interesting idea, but I don't really think that on-the-fly reconfiguration of another chip would be fast enough.

First, Windows isn't going away, if for no other reason, because the lion's share of businesses are locked into Office. So x86 isn't going away. Thus Intel is not dead-meat-waiting-to-happen. Wintel will eventually have a huge advantage in the tablet market (as well as desktop/laptop market) because of Office.

Second, if Intel didn't have something good coming for smartphones, would Nokia have tied itself so closely to Intel with Meego? I don't think so.