it ends up requiring segwit and still using backdoors to implement it (going soft using the anyoncanspend method for segwit has been admitted as a backdoor)
it also increases the TIER network strategy rather than a peer node network. and opens up more back doors.
having to 'opt-in' much like 'opting-in to use segwit keys does not 100% fix the promises that all these DCG partner bips keep promising to fix.
so although i like it because its another diverse node away from blockstream(core) domination. the gesture of fixing things is not a real fix.
it ends up leading to as i said more of a tier network.
differing UTXO sets.
different levels of validation nodes
some tx's seen by some nodes but not others..
it just makes a peer network ugly and non consensus and no longer existant...
which is the ultimate betrayal of the bitcoin ethos. because although is then more 'diverse' its a tier network so less decentralised.
plus it relies on back door implementations.
what needs to be done is an actual fix that even people with funds from the last 8 years can use without having to bait themselves into something that separates them from other nodes.(meaning a solution where all nodes see the same thing)
the real fix is to use real nodes consensus and just upgrade the network properly
Yeah Extension Block being based on SegWit doesn't sit well with me either. Who knows if it's just some kind of good cop bad cop routine.
But the Bitcoin well has been poisoned by BlockStream/Core so badly, anything that can immediately get the blockchain out of their death grip is a plus, hopefully after the great escape, there will be big diversity on implementations, then someone come up with a truly great and open solution from scratch.