A Powerbook G5, soon ?

According to Eweek, IBM would be about to deliver a G5 chip for laptops. Thanks to their PowerTune technology the electrical consumption of such chips would have been greatly alleviated.
Hence is Steve Jobs going to announce a Powerbook G5 soon?
Though this is something we could hope, we don't think this will prove true:
1) The Powerbook line has been renewed a (too) short time ago. Releasing much superior machines would result in many unsold units, and would make both clients and resellers angry. To announce a G5 Powerbook now, that would be available in September, would do pretty much the same.
2) IBM would claim they have a chip made for laptops. Yet is it necessarily a chip for a Powerbook ?
- One of the most important problems with the G5 currently, which is also its very strenght, is its frontbus that runs at half the speed of the CPU. On a Powerbook G5 2GHz, there would be a second, very complex chip (the chipset) whicch would work at 1 GHz. This chip dissipates heat, quite enough for having its own heat dissipator on G5 PowerMacs, using heatsinks technology. The energy used by this chip would sip a huge amount of energy from the battery.
- Other components associated to the G5 also dissipate much heat : the tension regulators, which are quite hugee. They must be cooled very efficiently, or tension would fluctuate and would involve problems for the CPU. If you take a good look on the photos of the G5 Dual 2.5 GHz Watercooling system (ICI you'll see they have their own heatsink and a HUGE radiator.
- Then you'll have to know what you call a portable computer. There indeed are portable PC which embark top of the notch CPUs, which don't work for more than 45mn and have the size and weight of two Powerbooks. Would you imagine Apple might release such a machine?
Now, it wouldn't be unbelievable that Apple great ingeniosity and skill could result into a line of portable G5s. Yet the obstacles are such, that we feel this is a prematurate hope. On the other hand, should IBM manage to master the 65nm die technology quick enough, it would be much easier to make a 1.8 or 2 GHz G5 that would require much less eenergy. The same thing would be requested for the chipset of course.