But designers tend to drastically underestimate discoverabity, particularly discoverabity by novice users. Many entire software companies exist purely because the market leader's product lacks discoverabity of a feature. Twitter, in some ways, exists because blogging software doesn't make it obvious that you can use their platforms for microblogging.
If you think of usability as a funnel, with discoverabity feeding into learnability feeding into usefulness feedin into ease of use feeding into efficiency feeding into fun, discoverabity is like your home page. It's where you have the biggest drop off of engagement typically, and it's where problems can absolutely make or break you.
Offscreen and corner gestures, while useful and efficient, are often so undiscoverable that they almost exist only for power users. On the lates Build and Analyze, Marco Arment said he has to include a button to show the side navigation on The Magazine because so many people have no idea they can swipe in from the left.
Apple typically gets this better than most companies, and will use a text button instead of a gesture because they know that even if 80% of users discover the gesture, they just can't rely on it for your core interaction. Because the 20% will just walk away and tell all their friends the product is crap.
My prediction is that over-reliance on gestures and hot corners will put a serious damper on Windows 8 and Windows Phone's network effects. It's good design for power users, but it's in no way universal design and for an OS universal design is a must.
I think you either missed "the importance of" or meant "overestimate".
It's that hardware feature that stands for 95% of the usability of the device.
Personally I think the product is incredible in many ways, and overall very overwhelming/impressive relative to my expectations, however to me it has two very core faults:
1. Performance. Many actions are perfectly fine, however v things like videos can be choppy, and even reading a Kindle book is choppy.
2. DPI of the screen. It's a nice screen, and the responsiveness even beats the iPad (incredibly), though after using new high-res displays, the Surface screen looks pixelated (primarily when rendering text).
A friend with $500-600 comes to you and asks what tablet he should get. Can you recommend a surface over the latest iPad? We can't say "oh, give it time, the ecosystem will get better, hardware will improve" -- your friend needs something today. (That's why it's underwhelming to me. It isn't a good deal against its competitors.)
I don't think there's an obvious answer to which tablet provides more value - it's very context-specific.
In the same way, if someone wants a top of the line tablet, it just seems obvious it's an iPad. That's what almost anyone with a tablet is going to have. It's just going to fit the expectation of what a tablet is for a naive user. Surface doesn't. Just like Linux on a desktop may be better in your context-specific suggestion, but it isn't what a naive user thinks a computer is.
Unless the friend also wants to pay me regularly to teach or tech support their computing device, in which case I might go outside the expected zone, but I've yet to find this to be the case.