Yes, the chances exists.
But my point is check if any of those dormant addresses from the past were generated using any of this methods, like brainwallets, sha-256 over a tx id, merkle root or block hashes.
It's just more a proof of concept than a "brute force" attempt.
SHA
brainwallet is easy, there 'were' about 2,000 addresses from the holy-grail 60gb dictionary that can be found that at one time were high-value, but sadly all those coins are now depleted, I have ran countless advanced brain-flayer on steroids using 100+GB dictionarys and frequently find new coin, but the money is always gone.
U must remember that those 2,000 golden addresses like sha(satoshi1), or sha(cat), they're all search 24/7 by bots and if somebody uses those addresses their money is swept in one second, just stating a fact,
Thus 'searching' address space using sha or any hash algo dictionary is a dead-end, then only real end is going after the high-value public-keys, using high speed advanced methods
***
If I didn't hate amazon so much, I would just setup up a huge AW2 system to search, but I hate giving them money, or using google, or facebook, don't want to feed the bastards
China has cheap, small light-weight GPU cards ( bitmain ) has some, once price goes to a few dollars we should be able to do 100's of billions of keys per second, using pollard-rho algos
***
None of this really matters, once BTC is cracked, they'll just bump from 256 to 512, and then the entire process just repeats; hamsters have been running this cage since cpu's were 1024 bytes of ram
**
U mention merkle, block hash, ... I tried all that stuff years ago, a few morons made keys using 'data', but nobody with real money backed his 'coin' using these techniques;
1.) hashed 'dictionarys' were the early easy gold to be picked off ground, think brainflayer 1.0
2.) using data to be hashed is an urban myth
3.) same for monkey at the keyboard the bible and unix-source will find you nothing, when hashed, I have ran all known text years ago through sha to generate priv's and found nada high value key, not even an empty one
The only real way is smart use of algo's ( pollard-rho, ... etc ), or study how the random keys were generated and use that for a basis of generating test keys for finding priv's
***
U can run gpu-flayers smart, using the 'algos', but the problem is you must let them run, and be very careful about the bloom filter that you put on the cards, as you don't want to deal with false-positives as it slows down the search
I used blooms on the gpu to find worthy keys, then use huge tries to see if they were every used, and if so then use high value tries to filter out the chaff, but when your doing 500 M-keys/sec, even a false positive of 0.0001 generates a lot of garbage to be analyzed
***
I looked at lbc, large bitcoin collider, and the blm filter he gives away and some of his writing and stuff, they're just doing a linear search, and to be honest they don't know what they're doing IMHO; doing a 1 to 10**77 linear search with a dog-shit bloom filter, which means they're just burning electricity; The real problem is in this game is get away from these 512mb toy blooms on brainflayer and jump up to 8gb blooms on on the 1080 class cards, then your good for 1 billion high value addresses and can search fast with low false positives
