is it faster if we use brute force attack on addresses in the puzzle by tools like bitbruteforce, plutus, btcbf... I mean these tools are generate random wallets (in 2^256 range) then check a list for matching an wallet with balance, which chance to success is much lower than 0. If now we use these tools but control the range to generate wallet only in 2^64 or 2^71-74... to find matching address in puzzle (the address list still can contain funded addresses).
I think this method will give more chance than kangaroo/bsgs or bitcrack, while solving puzzle they can find some funded addresses but still ignore them because it's not match with their target - addresses in puzzle only