Re: DECENTRALIZED crypto currency (including Bitcoin) is a delusion (any solutions?)
What are the 8 major flaws you perceive (keep it concise, please)? Then, we can try to guess what the key design innovations are. 
They are not exactly the same between PoW and DPoS even though I used the same numbering for both. I will not write the ones for DPoS at this time. For PoW, they are:
Quote from: AnonyMint's whitepaper
Satoshis Proof-of-Work
1. Permissionless free market to stand up a consensus ordering (i.e. mining) node. <--- not a flaw, but it is a flaw of DPoS
2. Power-law mining distribution is unstable trending toward winner-take-all devolution.
3. Transaction fees are not a free market because of the block size, regardless whether block size is fixed or variable.
4. No asynchronous concurrency of transaction ordering for enhanced scaling and throughput (although resilency and liveness are provided by miners competing to produce the next block). Only the node that produces the next block may add transactions during the block period.
5. Long range retroactive chain forks (aka long-con attack) if quantum computers appear.
6. 51% attack on minority blocks cant be objectively observed.
7. Orphans blocks. High transaction confirmation latency. Confirmation only probabilistic, never final. Burning must exceed double-spend value.
8. Wasteful and inefficient.
1. Permissionless free market to stand up a consensus ordering (i.e. mining) node. <--- not a flaw, but it is a flaw of DPoS
2. Power-law mining distribution is unstable trending toward winner-take-all devolution.
3. Transaction fees are not a free market because of the block size, regardless whether block size is fixed or variable.
4. No asynchronous concurrency of transaction ordering for enhanced scaling and throughput (although resilency and liveness are provided by miners competing to produce the next block). Only the node that produces the next block may add transactions during the block period.
5. Long range retroactive chain forks (aka long-con attack) if quantum computers appear.
6. 51% attack on minority blocks cant be objectively observed.
7. Orphans blocks. High transaction confirmation latency. Confirmation only probabilistic, never final. Burning must exceed double-spend value.
8. Wasteful and inefficient.
There is a clever twist in my design on Byzantine agreement. PoW is unbounded participation but never final, only probabilistic. Byzantine agreement is bounded participation and final. Both can be stalled indefinitely by 51% attack, and both have failure modes (loss of Nash equilibrium) with a 33% attack. My design combines both concepts but doesn't need PoW (formerly I had unprofitable PoW but hence replaced it with something better)! That is a very strong hint for you.