Analog sensing for keyboard switches has a lot of potential IMO.
Even if you are just having the keyboard send keystrokes to the computer in the normal way, with a good implementation of analog sensing, you can remove many of the compromises inherent in traditional switch designs. For example, you no longer have to balance your amount of over-travel for comfort against how much motion it takes the user to release and re-press a held key. You can do a clicky switch that double-taps even more easily than a linear switch. You can have the balance between initial press speed and resistance to accidental keypresses adjustable by the user or even application-sensitive, and making the key trigger sooner in the stroke will not carry any compromise in how far you have to lift your finger to release a key that's fully held down.
The main question will be how much of this potential Wooting is able to realize. Even if they half-ass it and don't get out of the mindset of thinking of activation as a point, there may still be advantages in durability and latency, as well as whatever specifically-analog stuff they manage to implement, so it'd still be kind of exciting-- but perhaps not as exciting as if they implement the switch more cleverly.