That's what I think, the kangaroo algorithm works, but you only get more efficiency with more computing power or a better implementation of secp256k1, there is no way to speed this up otherwise without making it less efficient.
So now that we know that gmaxwell created this puzzle. How should we continue? I think he should stop it.