Unfortunately, bitcoin has other severe design damages. It's code should be completely rewritten from scratch, preferably in Haskell programming language. Because bitcoin algorithms has business with huge amount of money, they must be formally verified, like seL4 microkernel.

So, that is HUGE amount of work, I can not do it myself alone. I may just write 'boundary' design conditions:

- mining amount of BTC, per unit time, should be proportional to CPU/GPU power, but in degree LESS than 1, till 90% BTC will be mined, algorithm must prohibit any parallelisation like ATI GPU does with current bitcoin algorithm

- distributed system should reject any possibility to group miners or peers in pools by design: they should be completely independent forever

- mining difficulty should be normalized to attract at least 1 000 000 miners world-wide, being on-line each day, on average

- transaction propagation should work in 'soft' real time conditions: we must guaranty that ANY transaction should be completed in 10 minutes, without paying fee, and if fee is 1% of transaction volume, we must guaranty that ANY transaction should be completed in 10 seconds

- if client wish, he could able to get possibility to de-anonymize, and to attract third-party that can roll back a transaction, if second-party is unhonest