Thanks for the thoughtful comments.
Yes, I agree there's lots of overlap in iOS app store. But suppose that just 10% of those 450,000 apps is useful. That's still a pretty big number. What percentage of Windows Store Apps is useful compared to the percentage of iOS apps that's useful? What actual number of Windows Store Apps are useful? What actual number of iOS apps?
Which apps matter will vary by user, of course. For some users, there will be a few Modern apps that matter enough to move someone to a Windows tablet. For others, the tablet experience simply isn't as important as highly mobile access to legacy applications, so they buy a Windows hybrid because they "get" a tablet in the deal but "need" a "real" computer. But on the whole, if the Modern UI and iOS were interchangeable as mobile experiences, wouldn't a few more Windows tablets have sold? Wouldn't users download more Windows Store apps than they do? If the number of apps isn't relevant, then I'd argue that device usage and app revenue provide at least a bit more insight. Apple happens to be pretty dominant in both of these metrics, which suggests people are in fact using iOS apps (and presumedly finding value in them) at a much higher rate.
Just one example-- suppose you're a basketball coach who uses a connected basketball to help high school players improve. I know of iOS apps that do a pretty good job collecting this data and making it actionable. I don't know of Windows Store apps that do the same thing. That's just a random, niche example-- but it's also an example of a distinctly tablet use case, and one that the iPad does better (as far as I know) than Windows. Suppose you run CAD programs, which is generally more of a desktop thing, but you prefer the way you interact with models in the iOS UI. That doesn't make the iPad CAD program more powerful, but it makes it useful-- and at least for some users, even more useful than a desktop version.
I'm emphasizing Windows app store vs. iOS app store here because in terms of the tablet experience, we shouldn't include the Surface Pro 3's desktop capabilities. That's a different use case, and not one that really competes directly with the iPad Air's use cases. If someone wants desktop capabilities, I have to wonder why they're looking at an iPad Air in the first place. It can be a productive device, including in some ways that overlap with PCs-- but the experience of using a PC is different than the experience of using an iPad. Similarly, the experience of using a MacBook Air and iPad is different than the experience of using a Surface. I think Surface Pro 3's strengths overlap more closely with the MacBook Air's than with any iPad's.
That brings me to your second point, about things that an iPad can't do but that a Surface can. You bring up good points. All of those examples are among the Surface Pro 3's strengths. They're why I said it works best as a laptop but is lacking as a tablet. It's also why I said Surface Pro 3 and iPad Air 2 will be compared too easily. They're different devices with different strengths. Just comparing the two of them as "tablets with large screens," as some publications already have, is significantly conflating some meaningful distinctions.
All that said, it's a good time to be a customer, given how many options we have. You want a cheap tablet? Great, there are Windows and Android options galore, and the iPad Mini is now reaching fairly affordable levels. Want a quality "pure" tablet? Great, you've iPads, Samsung devices, and so on. Want a 2-in-1 device? I think the app gap is revelant, but some of the Broadwell-class Windows 2-in-1s are going to be beautiful peices of hardware. I don't think there's "one device to rule them all," which is why I find the Surface vs. iPad comparison a bit suspect. Different devices are good in different situations-- and we have more devices from which to choose than ever!