We want to keep the EIP-1559 update rule consistent with the definition of how full a block is. To illustrate why this makes sense, let’s think about an example in a 2-resource design where the resource vector is (18M, 18M) and the limit is 36M. With multidim metering, the block utilization would be 50%, which means that we would be at target. In that case, we don’t really want to increase the base fee. If we were to increase the base fee and demand remained the same, we would see the block utilization drop below 50%. However, if we used the current formula for the base fee update, the block utilization would be 100% and the base fee would get the maximum increase.
Because they get more fees/MEV opportunities from them.
I don’t think they are comparable. With multidim metering, we can pack blocks more efficiently, which gives us a boost in available block space, independently of further increases in the gas limit. The idea is to align the block validity condition more accurately with the various resources’ safety limits.