The power of the finger

Well it seems Microsoft is really sold on the idea of creating touch sensitive interfaces. So much so that earlier this month(June)/the end of last month(May), they announced a new device known as the Surface Table, which basically begins to deliver on the idea that touch screens are going to be the interface of the future. Well todays post/video shows how they're adapting it current products in this case a laptop, meaning the future maybe coming sooner than we thought

So now that you've gotten the basic press pitch, here is my two cents. The technological approach Microsoft is taking to implement this technology seems to be pretty solid, however, I have some concerns:

1. Will these devices deliver good image quality? Flat display technology is quickly becoming a mature technology but it does have a rather rough track record with behind the glass sensors. When it does work it give the screen a slightly grainy look (a la Tablet PC) so the question is can Multi-touch interfaces integrate well into portable devices, such as laptops , and still retail the great display quality we all know and love?
2. Are people really going to want to reach out can touch their display interface? Now this just maybe because I'm pretty well fixed into the input and display paradigm, but think about it are you really going want reach and touch your screen all the time? If not then you're at least going to want a keyboard like device in the very least and possibly a mouse just because it seems more of a natural feel. Personally I think the concept works better as a table since it puts the interface at your finger tips rather than your finger tips at the interface.
3. Isn't this just an interface device and not entirely a mouse replacement? The main issue I have with the notion that touch interfaces will replace our good old little pointers is that they're inherently less accurate meaning they not good for very precise pointing. If you want to know why that's important ask any gamer or computer graphic artist. So have we just found a fancier pointer or is this just an interface problem no ones really addressed yet?