I want to bring another point into the latency debate.
We cannot treat L1/L2’s in a vacuum and need to consider competing trading venues in the model. Let’s only consider trading as a use-case (71% of validator payments come from trades)
- Latency means slower price discovery
- i.e. Other venues will determine the price and have higher volume (due to how MMers work)
- LPs in the slower venue will leak value to toxic traders (maybe innovations like McAMM might reduce the leak)
- LPs prefer to go to higher volume venues because they have more fees and lesser value leak reducing the liquidity in slower venue.
L1/L2 designers will prefer to have more trading volume since it increases the value of their network. More transaction fees and more MEV to be captured and distributed.