Though there are no specific or official details on any of these, a lot of them can be assumed based on previous hardware.

The likelihood of Nokia offering the new device without a hardware keyboard is extremely low. The virtual input methods became filled with regressions after the release of the N810, and improvements have not been forthcoming. The state of virtual input is reasonably strong evidence of Nokia not considering it an important feature for future devices. Besides, the N810's implementation has proven to be a well-received decision that they'd be unlikely to go back on.

Now, the usability of any hardware keyboard they design is certainly suspect.

It'd be safe to assume that neither of the first three will be likely to deviate overly much from existing hardware. Quality of construction, well, whatever's planned can always be undermined by unforeseen manufacturing defects.

This one is pretty safe to extrapolate to a reasonable ballpark based on current hardware and some basic information about the OMAP3. Best-case, it'll be significantly better than current hardware (as the OMAP3 at idle uses significantly less power than OMAP2 at idle—crazy, right?), worst-case, it'll be marginally worse (the maximum consumption of the OMAP3 is a bit higher than OMAP2, especially when you factor in the PowerVR—so if you're listening to music playing Quake III, you're likely to get less battery life ).

MicroUSB with OTG support are extremely safe assumptions. Whether it'll charge over USB, well, does the N97?