I loved my MDs and Hi-MDs, but they had so many frills. All the frills. That was part of why I loved them!
Redkey
The PlayStation 1 had a copy protection system that measured physical properties of the disc which couldn't be replicated by normal CD writers. There were a few ways to get around this, but to be able to put a burned CD into your console and boot directly from it into the game (as usual) required the installation of a fairly complex mod chip. A lot of people alternatively used the "swap trick", which is how I used to play my imported original games.
The DreamCast's copy protection was heavily reliant on using dual-layer GD-ROM discs rather than regular CDs, even though they look the same to the naked eye. There were other checks in place as well, but simply using GD-ROMs was pretty effective in and of itself.
Unfortunately, Sega also added support for a thing called "MIL-CD" to the DreamCast. MIL-CD was intended to allow regular music CDs to include interactive multimedia components when played on the console. However, MIL-CD was supported for otherwise completely standard CDs, including burned CDs, and had no copy protection, because Sega wanted to make it as easy as possible for other companies to make MIL-CDs, so the format could spread and hopefully become popular. Someone found a way to "break out" of the MIL-CD system and take over the console to run arbitrary code like a regular, officially released game, and that was the end of DreamCast's copy protection. People couldn't just copy an original game disc 1:1 and have it work; some work had to be done on the game to put it on a burned CD and still have it run (sometimes quite a lot of work, actually), but no console modification was needed. Anyone with a DreamCast relased before Sega patched this issue (which seems to be most of them) can simply burn a CD and play it on their console, provided they can get a cracked copy of the game.
I think it depends a lot on a person's individual knowledge. If you keep studying far enough away from your main area of expertise, there'll still be some point where you stop and have to blindly accept that something "just works", but it will no longer feel like that's what your main field is based upon.
Imagine a chef. You can be an OK chef just by memorizing facts and getting a "feel" for how recipes work. Many chefs study chemistry to better understand how various cooking/baking processes work. A few might even get into the physics underlying the chemical reactions just to satisfy curiosity. But you don't need to keep going into subatomic particles to have lost the feeling that cooking is based on mysterious unknowns.
For my personal interest, I've learned about compilers, machine code, microcode and CPU design, down to transistor-based logic. Most of this isn't directly applicable to modern programming, and my knowledge still ends at a certain point, but programming itself no longer feels like it's built on a mystery.
I don't recommend that every programmer go to this extreme, but we don't have to feel that our work is based on "magic smoke" if we really don't want to.
ADDED: If anyone's curious, I highly recommend Ben Eater's YouTube videos about "Building an 8-bit breadboard computer!" It's a playlist/course that covers pretty much everything starting from an overview of oscillators and logic gates, and ending with a simple but functional computer, including a CPU core built out of discrete components. He uses a lot of ICs, but he usually explains what circuits they contain, in isolation, before he adds them to the CPU. He does a great job of covering the important points, and tying them together well.
I was so triggered by the conversion from char-to-int-to-string-to-packedint that I had to write a bitwise version that just does char-to-packedint (and back again), with bitwise operators.
https://pastebin.com/V2An9Xva
As others have pointed out, there are probably better options for doing this today in most real-life situations, but it might make sense on old low-spec systems if not for all the intermediate conversion steps, which is why I wrote this.