The Sony 70-200mm F2.8 does indeed cost about the same as the image-stabilised Canon and Nikon equivalents (although rather more than the non-stabilised Canon). But the big advantage of in-body IS is that all lenses are stabilised, so it competes with the Tamron and Sony lenses on a much more equal footing. Compared to these two, it's double the price or more, and you don't appear to gain much in functionality for your money. That's the comparison I will necessarily have to make in reviewing the lens; what does $1800 for the Sony buy you over $800 for the Sigma?

Re: the Panasonic 14-150mm, we'd have no problem getting a sample to review, but we have to decide whether it's of sufficient interest and benefit to our readership as a whole to commit the resources. That's not trivial by the way; that four page Canon 18-200 review is based on no fewer than 790 test images in the studio, backed up by over 500 shots simply using the lens and seeing how it performs in the real world. And no we don't keep the lenses, they (naturally) get returned to the manufacturer or distributer who provided them.

As for whether in-lens IS has any negative impact on image quality, we already have plenty of data to suggest it makes very little difference at all. It's easy to make this comparison in the lens review widget, e.g. Nikon 70-200 VR vs Tamron, Canon 70-200 IS vs Sigma, Canon 18-55 IS vs Pentax, Nikon 18-55 VR vs Olympus 14-42, Canon 17-85 IS vs Sony 16-105. Overall this shows interesting differences, but nothing which can be ascribed to IS. Of course we also know Canon's 18-55 IS is much better than the older version, but there's no way that could be used to argue that addition of in-lens IS improves optical quality.

'The Canon non-IS 70-200mm appears to have better optical performance than the stabilised version'

Only half-true I'm afraid - the non-IS F2.8 may be slightly sharper, but between the F4 lenses the situation is reversed. Any result in the Tamron 18-250 vs 18-270 comparison would equally be just another isolated data point, and not indicative of any general rule.

I fully understand that there's a desire in certain quarters to 'prove' once and for all that sensor-based image stabilisation is better/worse than lens-based, but as someone who's used both systems extensively, I can only say that the answer is a resounding 'it depends'. Both work, both have their advantages and disadvantages.

The moving group in IS lenses may have a slight negative effect on IQ, but equally the fact that a moving sensor necessarily tracks across areas the lens's image circle with slightly differing resolution and geometric distortion isn't going to help. In both cases the limiting factor is more likely the control system, and how accurately it can compensate for shake given the inevitable processing lag and the physical inertia of the sensor or IS group. Both systems let you get better images than without IS, but neither will be quite as good as using a tripod.