Subchains and other applications of "weak blocks"

[Matthew Light]

"The use of subchains also diverts fee revenue towards network hash power rather than dripping it out of the system to pay for orphaned blocks."

I can't see orphaned blocks as ever constituting more than 20-30% of block solutions in any realistic mining regime.

Network security versus insecurity isn't going to be make-or-break with a ~20-30% delta, IMO, unless I'm missing something.

That said, reducing orphans is a good idea, as is coordinating the block-building activity via sub-chains, as it does enable much larger blocks to be built for higher network throughput.

 

[Joseph Woolfridge]

Looks like nullc aka Gregory Maxwell is on a posting spree of late. He goes on a tangent with someone about Peter R's work, even though the guy didn't mention subchains or @Peter R .

 

[YarkoL]

Outside of the assumptions, the arguments presented also do not hold since miners can coordinate to only include transactions in blocks once they are already well known, eliminating size related orphaning completely.

Actually this point disturbs me. SPV mining etc. already go long way towards headers-only block propagation and then the supply-demand argument seems to break up.
@Peter R , can you ease my mind on this?