After reading a lot here, at least for me, I can see that the idea behind prefix analysis isn't going to work (sad, I was trying to come with something too). Because the X point is passed through SHA256 and RIPEMD160 and the hash result is very random. So no pattern can be found or a high probability that the key is on that range. If this kind of analysis were effective, why haven't people who solved "higher" puzzles continued to solve subsequent ones? Because it was solved with brute force and a lot of CPU/GPU power using pure brute force / BSGS / Kangaroo.

With puzzle 66, I was very close with a prefix I found. But I also found a lot of others prefixes. How do I know which one to follow? This is the main problem.

But, this is where the fun continues. I will continue to run my scripts. Maybe I'll win the lotto and find the solution. hehehehe!