I am a Software Engineer and I spent a couple of weeks reviewing popular crypto coins source code. I am not going to comment much on the cryptography aspect of, as I approached it as just another piece of software.
Bitcoin was pretty solid. I remember when it first appeared I looked at it and concluded it to be a "great PoC". I totally expected the humanity to improve on its design in the future. It did not. There are a couple of things that irks me as a Software Engineer.
The first is its single point of failure. Everything hinges on just a singular operation - modular inversion - to be slow. If someone figures a fast way to do it, they'd be able to crack private keys in no time. And while this single point has been holding strong for the past several decades, an Engineer in me is still not comfortable with it. It is just a single tiny operation and the entire Bitcoin ecosystem depends on it.
The second is the arbitrary drop of encryption strength. Most of the cryptography is done using 256 bit. But when it gets to the blockchain, it drops to just 160 bits. That is still a lot. But it is also a lot less. 96 bits is 30 orders of magnitude! As an engineer, I do not see any reason to do that. Well, you save about 8 bytes from every public address on the blockchain. Does it really warrants dropping the cyrptography strength by so much?
I expected the follow up coins to do better.
The next big one is the Ethereum. First, it is based on the exact same curve as Bitcoin. But instead of double-hashing the point's coordinates, it only single-hashes it. And the way it chose to represent the point before hashing is distinct, but oddly compatible with Bitcoin's. In a 65-bytes buffer you can write a sign marker, x and y coordinates and the first 33 bytes will be the Bitcoin's input buffer, and bytes from second to the end will be Ethererum's input buffer. As an Engineer, I do not believe in coincidences like this. The Ethereum coin is designed such, that it is easy to crack together with Bitcoin.
As a result, the same cryptographic curve hanging on a single point of failure is populated with "points of interest" by two independent coins, and designed such that it is relatively easy to crack the two together.
Does not look reassuring. So I decided to look at major modern coins on a different curve. Enter Solana. I expected it to be better. But it turned into a deep rabbit hole and deserves a separate blog post on its own.