Your reasoning is actually very logical, and you did a good job at backing it up.
It seems in your case it would be very difficult to find pure examples of invention. If we keep looking at the predecessors of all inventions we eventually get to the natural elements of the universe.
Which is why I consider Apples products to be inventions. You know, it's not like they took the parts of that 1998 MP3 player in your post, and rearranged the parts to make a better MP3 player. They introduced entirely new parts to the system that were never a part of MP3 players before. Isn't that invention? The click-wheel was totally original. Isn't that invention?
When you get into a mature market, such as what the computer hardware business currently is and has been for years, finding an example of pure, never before seen tech is a rare thing. Processors are like the internal combustion engine in a way. There are a thousand ways you can make one. You can make it smaller, more efficient, add a couple of pieces here and there, but no matter what you do, it's all just a tweak on what first showed up back during the early 20th century. You could make the smallest, most fuel efficient engine the world has ever seen, and you wouldn't be able to patent it because it's still the internal combustion engine (though as an aside, you can patent the manufacturing process if its unique enough).
Apple hardware is the same way. They can design really tiny computers that pack pack a huge amount of punch for their size, but they're still working from the same blueprints provided them by Intel and ARM Holdings, same as everyone else. The A7 isn't much different than a Snapdragon or a Tegra. They're all ARM chips running off the same instruction set. The only difference is how they're put together.
It's the reason why patent disputes are over tiny, incidental things these days. Computer hardware is standardized to hell and back, and rearranging the parts doesn't necessarily make it something brand new, just potentially more efficient. Until we start moving away from silicon into, say, quantum computing, there isn't anything shockingly new that can be done. It's all iterative by this point. Everyone takes the same thing, and makes it smaller, faster, lighter, and better.
Everything from calculators, to MP3 players, to smartwatches, to smartphones, to tablets, to laptops, to desktops, to servers, they're all basically the same thing, using the same tech in different ways.
Oh, and this kinda shocked me, but the click-wheel was developed for Apple by Synaptics. So even there, they made the design, but they didn't invent the tech behind it.