Re: Killer App & Price
When they're going for the consumer market, a lot of those apps aren't worth the effort of Apple writing them. Someone else writing them, sure, but not for Apple to do the work before launch as an advertisement. For example:
"Link up with Boeing or Airbus and have an app that shows all of the parts & part numbers of, say, a fuselage listed in front of you as you look at it, along with possible diagnostics, numbers of spares in local warehouse?"
To the typical user, this is something they'll think about for about ten seconds: "Cool, an aircraft mechanic is using these", then ignore. It could be useful to the aircraft maintenance industry, but there's no reason for Apple to write that instead of providing encouragement and support to that industry to do it.
"Or an app for a surgeon that lists detailed anatomy of a patient opened up for surgery along with possible noteworthy points about what is on screen?"
I don't think that's going to work with current technology. It isn't a part that has a deterministic appearance which can be identified quickly. And if anything goes wrong, the press will be a lot worse than any benefit from advertising it in the first place.
"Or an app that shows you, step by step iFixit-style how to repair, say, a bike, a dishwasher and the like?"
If they could get this one, that would actually be a great advertisement. The trouble is that they can't, nobody can, and I think you know that. They don't have the software to identify automatically which model of bicycle I have when that model could easily be decades old from any country and then identify the problem with it from a glance, then automatically provide me a useful solution when one might not exist. They don't have anywhere near the staff necessary to manually accumulate that data either. There is a reason why there are only iFixit guides for the most common devices. If you bought a cheap phone instead of a flagship, you likely have to do the dismantling yourself to figure out what's in there and whether you can fix it.
"Or an app that shows you how to cook an elaborate meal along with hints, suggestions and instructions while you are doing it'"
I suppose that one is a bit more realistic, but it's not really any more than a video that starts and stops. Watching you do it and determining whether you've made mistakes is trickier and not very useful, since anyone who wants to do this can probably follow along with a video unless they're intentionally trying to test the error recognition system.
"Or an app that explains how to solve, say, the maths problem in front of you using a variety of different ways?"
Why do they need this? There are already programs out there for teaching and solving mathematics and they don't really benefit from AR.
"Or an app for farmers looking to buy cattle at a cattle mart: display details of the animal in front of you, [...]"
Again, something a small set of users would use and everyone else wouldn't care about, and in order to build it, they would need to get access to all that data. Every place with livestock displayed would need to put those details somewhere the app could get it and tag things so the equipment could pick it up. I'm sure there are multiple competing databases where some of that is stored while some others simply write it down or have someone tell you with words. Apple trying to produce an iTunes livestock store and get everyone to use it doesn't seem worth it.