When I switched to Windows, I learned Visual Studio and C# and built several customized tools that coworkers could only use on Windows. I also recommended Windows machines to family and friends, who almost all went with my recommendation.
When OSX came out, I and many other developers switched to it. Finally, a serious OS from Apple. When we did, we pulled lots of families and friends with us, and we made a lot more things (open source installers, for example) available for Mac users.
If Apple succeeds at restricting Macs to people who want lightweight, fashionable, "Apple lifestyle" accessories and need confusing things like folder hierarchies removed from the OS, many of us will move again and take those IT decisions we influence with us.
If I get off the Mac, for example, I will quickly forget how to do things I need to do to provide tech support to Mac-using family and friends, so I'll probably begin recommending Chromebooks to many of them, shifting some of my support away from their individual machines and onto web portals and web apps I build for them on my own servers.
And I went with iPhone because I was a Mac user, and I figured I would get some benefits from using the two together, which I did. If I'm no longer a Mac user, those benefits vanish, and I'll probably switch to an Android phone with an audio plug. (Maybe we developers are just a vocal minority, but we'll probably get pretty vocal to others about why we're leaving Apple if we do.)
Every year they seem to continue to introduce (via neglect/decisions/priorities) more reasons to pause at purchase time. And, recommendation time.
In most families and circles people go to the person in the group who knows the most about tech, usually professional user. This person will often recommend products they themselves value and are familiar with. Obviously I can't prove this effect was partly responsible for Apple's come back, but it's certainly true in all the social circles I'm a part of and I think a decent hypothesis.
I don't think anyone is asking for your number two reference. It's more that people simply want their use cases considered. In many cases they are not. You can argue that the key changes are not crucial, and that the USB ports are future looking. But what about the lack of 32gb ram? The lack of new processors? What about the sdcard reader (yes, consumers mostly use phones now and don't need it, but professional cameras are going to be using sdcards for the foreseeable future)?