No, I haven't actually seen one yet. But the Leica M9 was announced a few days ago, and it's said to be shipping beginning in "mid-September." It's billed as the worlds smallest and lightest full-frame (18mp) digital camera, and indeed the body is almost unchanged from that of the M8. At $6995 US, it's pricey... but in the same range as top-end pro full frame DSLRs.
Back in August Leica invited a handful of photographers to Solms and gave them an advance look at the M9, a few days to use the camera with a variety of lenses as well as tour the factory. A few of those advance look reviews have been posted online in the past few days, and I especially recommend the ones by Michael Reichmann of Luminous Landscape at http://www.luminous-landscape.com/reviews/cameras/m9-first.shtml or Erwin Puts at http://www.imx.nl/photo/leica/camera/page155/page155.html and there's also one at dpreview.com .
It's interesting that Ken Rockwell has also posted one of his typically heavily opinionated "reviews" on the M9... but if you read carefully, he hasn't actually had his hands on one yet, he's just ordered his from the store the same way you or I would need to. That's right, you're reading correctly... he's making numerous strong statements on what he likes about a camera he hasn't even seen yet. While there's often some good information buried in his writing, it's most evident to those who already know as much about the gear in question as he does, and there's also often serious factual error and an over-aggressive and over-simplified black-and-white worldview that grates on my nerves. This latest stunt has cost him severely in credibility, in my eyes. Perhaps even more damaging is the fact that Leica apparently did not include him on the invitation list for the August event... an action which speaks volumes.
I found something far more useful in the process of looking at the Erwin Puts review. In an earlier post http://www.imx.nl/photo/technique/page153/page153.html he does a head-to-head comparison of actual prints from film run through an M7, and digital capture from an M8... with the same lens on each camera. This is what Puts excels at, the detail-freaky test chart resolution comparisons. And he's made a very important point: Most film-digital comparisons are actually comparing digital scans of film to digital capture in camera, which really is a comparison of scanners vs. cameras and not film vs digital. Thus, most film-digital comparisons are fatally flawed.
Here, Puts makes a fairly good case that looking at actual prints, the M7 still outperforms the newer digital M8 or M8.2.... although not by all that much. And he presumably did his test shots on tripods in the studio, which is not usually the way one uses a Leica M. I'm also not so sure this would hold up with most other camera brands, because Leica glass is so good that it exceeds the resolution capability of the camera in many cases. With most cameras, a slight resolution advantage of one media over another will be limited by the lens, by the mirror-slap vibration of an SLR, or by shortfalls in technique.
I find this of interest because it parallels my own experience. The M8 is good, and the M9 is expected to be slightly better. But the M8 isn't quite as sharp... actually, it's more about acutance than sharpness... as some of the things I've done on film with a 10-year old M6. I'd compare the look of the M8 to film processed in a finer grain developer like XTOL, which gives a creamy smooth look; but a lot of my M6 work is processed in Rodinal or HC-110, which give a grainier and higher acutance "edge" to a properly executed image.
But that raises another question, one not addressed by Puts... and this again is typical of his detail oriented worldview, he's better at the mechanistic things than he is at the big picture. But considering what a Leica is used for... fast spontaneous capture of Cartier-Bressons's "decisive moment" ... If it's ultimate technical image quality I'm after, rather than worry about an incremental difference visible only to the trained viewer, I'm more likely to skip the rangefinder (or the DSLR) and use medium format instead. Indeed, this is what I've been doing when I'm shooting under controlled conditions, on a tripod with studio flash. The big negative is in a league of it's own.
Thus, for on location shooting where big cameras and tripods are not the best choice, the digital Leica M offers a lot of advantages. The M8 is practically all I've used for location work this summer, and the M9 addresses most of the relatively minor annoyances I've encountered with the M8.
If I were willing to spring big bucks for an M9 (which I'm currently not), the things I'd be excited about would include the allegedly improved high ISO performance, the absence of a crop factor, and the relationship of the lenses to the brightline frames in the rangefinder window. With the older M bodies, with the standard 0.72x rangefinder magnification, the frames for 35mm and 50mm were well suited to in camera editing. That is, they're big enough to be able to see the image effectively, yet there's enough space around the frame to see what's being edited out, and to make conscious decisions about composition. This is one of the advantages of the rangefinder, one of the reasons "seeing" is different than through the dark tunnel of the SLR.
But with the M8, I find that only the 35mm frame (46mm equivalent with the 1.3x crop factor) allows this optimal composition. The 28mm frame (37mm equivalent) is out in the edges of the window, and not much is visible around it. The 50mm frame (67mm equivalent) is so small that it's hard to work with, and the 75mm and 90mm frames are essentially useless because they're so tiny. And since fast 50mm lenses are among my favorites on the Leica M, I find this frustrating.
The 0.68x magnification of the M8 (and reportedly also of the M9) also complicates focus a little. I initially scoffed at early reports of focus problems with the M8, since with the film Leica M bodies I find focus accuracy to be excellent, and one of the great advantages of those bodies especially in low light. But indeed, I've learned that extra care with focus is needed with the digital M bodies, especially with 50mm and up lenses. Erwin Puts attributes this partially to a narrower zone of high sharpness with digital sensors than with film (I'll take his word for it), but it's also because of the lower viewfinder magnification.
These issues can be addressed with addition of a screw-in 1.25x magnifier which threads into the eyehole on the back of the M8 or M9, and which brings the viewfinder magnification up to the same 0.85x that was offered on one version the M6TTL. I may very possibly try this, especially since my M6 is an 0.85x model and I'm already accustomed to that. Still, it's annoying to need to spend another $300 on top of an already expensive camera.
Finally, the lenses: With the M8, Leica made a big deal out of six-bit coding. With the M9 there is reportedly also the ability to set the lens ID manually via the menu, which makes it easier to use older (and much less expensive) lenses.
There's some attraction to the latest aspherical Leica glass, which is amazingly sharp and with high microcontrast, but also very expensive. However it's important to keep in mind that Cartier-Bresson, Riboud, Sieff, Gibson, and many other well known Leica M shooters have been creating exquisite work for many decades with lenses that, on a test bench, would not compare well with the latest versions. That didn't prevent them from creating all those classic photos; and I do not recall most of those photos as less than sharp.
And for that matter, they created them with older Leica M film bodies that are available used for a fraction (10-20%) of the cost of an M9, and which typically are still in mechanically perfect condition even after all those years. The classic versions of these cameras can last a lifetime.