GP is saying that if the [private] key is password protected, it can be brute forced. Surely if a human is able to type in the password to the private key, you can bute force that password relatively easily?
Of course, there are many other factors that can significantly decrease or break the security of AES.
TL;DR xkcd.com/538
"Easily" conceptually, yes, but not computationally.
>What does "check payment and receive keys" really do?
Presumably it asks the server whether the payment was received, and if the answer is yes, the server sends it the key.
Maybe malware operators aren't that experienced, who knows.
There was an early version that would generate the private key locally, which was discovered by Symantec (the original blog post isn't loading, but you can see it at https://archive.is/1AGHG) For an analysis of a more recent version that fixed this see http://www.secureworks.com/cyber-threat-intelligence/threats...
A TorLocker had flaws that allowed 70% of infections to be recovered, see https://securelist.com/blog/research/69481/a-flawed-ransomwa..., although they didn't explain what flaws there were.
What does "check payment and receive keys" really do? Does it check the Bitcoin block chain to see if payment has been made, then calculate the decryption key? Is this entirely blockchain-based?