Hard to say without knowing the counter factual, e.g. if SegWit was deployed with no discount and the witness data included in the 1 MB limit.
We don't know what part of that 50℅ was due to slower sync vs. e.g. just more user friendly wallets (that don't use a full node).
Without knowing which of these nodes actually controlled funds vs just astroturfing it's not even clear if the trend is really down.
Replies (7)
There's also the UTXO set growth to consider, a problem that takes a bit longer to manifest than block size. SegWit improved the incentive there.
Segwit did improve quite a few things.
That fix doesn't require a block size increase.
Without the discount it would still be cheaper to create a new change output than to spend an existing UTXO.
A discount with a 4x decrease for non-witness stuff would also work. That would have been a "bit" controversial though.
SegWit was not the only solution proposal for transaction malleability; remember BIP62 with Pieter Wuille’s proposed tweaks to the BIP in July 2014.
SegWit inadvertently prevented covert ASICBoost.
Incentives for UXTO consolidation could exist without SegWit discount; transaction byte size fee payment encourage already UXTO consolidation.
It kinda still is, thanks to the block size increase.
Without the block size increase, the discount could apply to all transactions
BIP62 is partially implemented, but the remainder is withdrawn. Afaik it wouldn't be enough to be safe for Lightning, but you'd have to ask (Sipa on) the Bitcoin Stack Exchangs.
