I am not a leet hax0r :(
Re-reading my posts, I'm not sure I've explained myself clearly enough, so I will attempt to do so now.
When I talk about things being a matter of 'implementation', I am by no means trying to belittle them or their importance to a product like the iPad. To paraphrase Edison, products are 1% inspiration, 99% implementation. People too often focus on and judge ideas, without realising they are far easier, and far less important, than implementation. The idea provides the hook that gets people interested, but the quality of the implementation is what determines its success or failure, and it's much harder to get right.
What is the essential idea behind a tablet computer? I would define it as a being a handheld device, where the display and all other components are contained within a single tablet/slate shaped unit, and where input is provided through the display, rather than through a physical keyboard or some other peripheral.
Nothing in that definition makes mention of the software the device runs, the quality of resolution of the display, the quality of the input. It is possible for all these things to change, without affecting the essential 'tabletness' of a device. That does not mean they are unimportant, in fact they are the most important aspects of the device.
For example, consider search engines. The basic idea of a web search engine is simple. You have a site that lets you enter some keywords, then returns a list of resources relevant to those keywords. In the nineties, search engines generally sucked. They were slow, returned shitty results, and made no money.
Then Google came along. They didn't change the basic idea of search at all. You inputed keywords, you got results back. But their implementation was better, massively better, in every way than all their predecessors. They used page-rank to return genuinely relevant results, huge data-centres to index the whole web and return results quickly, and made money via sponsored keyword advertising. As a result they became hugely successful.
Historically people have been too quick to dismiss an entire idea, such as tablets or VR, on the basis of problems with the implementation of that idea. When most people dismissed tablets, they didn't say 'tablets with desktop operating systems will never work', or 'tablets without capacitive touchscreens will never work', they said 'tablets haven't been successful, therefore people don't want tablet computers'. They were wrong.
I think, and hope, they are wrong about VR as well. The basic idea of VR is a head mounted display that takes over your entire view to create a fully immersive display. By itself, that idea is not a route to success. As history has proved, it's entirely possible to produce a shitty implementation VR, where the problems Carmack enumerates mean the end result sucks. But if someone were to come along and do a decent implementation, using the right hardware and software to createa truly compelling experience, then I don't think it would be a flop, I think it could be very successful.