We actually don't understand technology as well as we think we do. Well, as a society we might, but the average reader actually just barely grasps the tiniest bit of it. Extrapolate further into the future to approach Clarke's "Sufficiently advanced technology," and it gets even harder. What we actually have, as a society, is great faith in technology. We're willing to just trust that technology does what the scientists and engineers say it does. This can be effective for simpler things, but as you approach Clarke's sufficiently advanced technology, it helps less and less.
Take, for example, the iPhone. I can guarantee you that you don't truly understand how it works. Nobody does. Too many people were involved in that project. Some understand some small parts of it. Many understand a higher level view of it, but nobody really understands it as a whole.
Let's go smaller. On a single chip in the iPhone, there are millions of transistors. Transistors are basically 2 diodes smooshed together (handwaving a ton!). Diodes permit current to travel one direction, and not the other. Great, I just explained enough technology to use diodes in a story. What if I put too much "reverse" voltage on it? In electrical engineering, we call the resulting behavior "avalanche breakdown." It's an otherwise unexpected and very sudden increase in current going through the diode the wrong way (the direction current shouldn't be able to go). You can accept this on faith, as most do. Or we can go into why a diode works, and cover the quantum mechanical reasons for avalanche breakdown. They're really fascinating topics, but unless you have some background in the topic, you really just have to trust me when I say they work that way. And if your technological story ever resolves a crisis by 'reversing the polarity,' this is the kind of real life effects you would have to consider!
Let's go even simpler: solder traces. Solder is a low-melting point metal which is used to connect chips to circuit boards. In my story, the villain had amassed an unstoppable arsenal of doomsday weapons. At the darkest hour, the villain hits the button, but nothing happens. Deep inside each weapon, the hero had catalyzed metallic crystals to form, creating short circuits. Sound fantastical? In 2003 it would have sounded like a movie plot.
Then the EU passed RoHS, an act of legislation preventing the use of hazardous materials such as lead in electrical components. Solder was partially made of lead in those days, so they had to change its composition. Unfortunately, the new composition tended to cause tin to crystallized out into "tin whiskers," which reached across chips and shorted them out. It caused all sorts of failures, including a false alarm at a nuclear power plant!
The point of that example is that the definition of what technology we will take "on faith" is constantly changing. Ten years ago, the idea of crystals shorting out our electronic circuits would have been one of those "oh yeah, I'll trust you. You're the author, it's your story" type of plot devices. Five years ago, it was considered to be science fact. If the amount of faith we are willing to give technology changes that fast, you can see why technology starts to receive a magical treatment 100 or 1000 years from now in stories.
As for your second and third arguments, I agree that technological limitations are interesting. Just take a look at Apollo 13, which was a fascinating exercise in what could be done within limitations. However, I'm not convinced they're easier to work with than magical limitations. Technological limitations have a tendency to be very pass/fail. Either the whole idea works perfectly, or it fails catastrophically. Neither of these make for all that interesting of stories. The interesting stuff appears when the idea works just enough that the characters in the story feel some desire to expend effort to make it work all the way. It turns out that sitting in that narrow band between pass and fail is hard. It's much easier to do with magic because you can have looser definitions. And do remember that every one of those limitations must be understood (see the first rule).
As an example, consider AGIs: Artifical General Intelligences. They seem like a natural extrapolation of our current AIs. However, it can be tricky. Many of the mathematical tools you use to extrapolate from AI to AGI break down in funny places. I have a long history on this forum of abusing Godel's Incompleteness Theorem to poke at issues that arise when you naively extrapolate in this way. However, explaining this theorem is not easy. In fact Douglas Hofstadter spent an entire book on the topic, Godel Escher Bach. The limits put in place by the incompleteness theorems are not trivial to understand, so may call for "on faith" arguments which are bad if you're applying Sanderson's Law.
In all, the true complexity of technology is oft stranger than fiction, and its limiting cases can be positively bizare. Consider the sound barrier. As you pass through the sound barrier, you pass across an asymptotic in the velocity/pressure curve. Once on the other side, pressure actually has the opposite effect intended. This causes control surfaces to behave backwards once you cross the sound barrier. Modern supersonic aircraft handle this in their fly-by wire system, but pioneers such as Chuck Yeager had to address it manually! (Just one more example of why Chuck Yeager was such a bad ass!)
Well, almost. I'll have to be straight with you: all of that stuff about the sound barrier was made up. It was actually a movie plot for a movie released in the era of Chuck Yeager's famous flight, as recounted in his autobiography (EDIT: according to a4android, the film was The Sound Barrier. Thanks!). When asked about whether this effect occurred, he simply stated that, had it actually happened that way, he'd have been dead. If I managed to fool you with the previous paragraph, it goes to show just how far our faith in science and technology goes. (Don't worry, all of the other examples I chose in this answer are indeed real life examples. That was the only fake one)
The actual effect Yeager faced was that, as he crossed the sound barrier, he lost all elevator control. He could not control his pitch up and down. This actually stalled efforts to break the sound barrier for a bit while they tried to figure out what happened. The actual cause was the shockwave from the tip of his nose passed across a hinge on the elevator and created enough pressure to prevent it from moving properly. Once he punched through the barrier, the shockwave moved to a different place, and he regained control.
Oh, and control reversal is a real phenomena, but it occurs for completely different reasons. The Supermarine Spitfire had an issue with this due to insufficient torsional stiffness in the wing. In some of its extreme dives, the air forces on the wings were sufficient that the ailerons stopped functioning properly. If you commanded the ailerons to go one way, they'd actually end up twisting the wing the other way and the net effect was that the plane would roll in the opposite direction from what the pilot intended! So maybe the truth really is stranger than fiction.
