Vitalik Buterin: "Reducing L1 slot time" is worth discussing

Bankless founder David Hoffman said in a post on X that, in a sense, the rollup-centric roadmap has been "completed", and progress has been made so far, including the launch of 4844, DA, and inexpensive L2. In the future, the primary optimization focus is L1.
Vitalik Buterin responded: "I don't think this is a priority issue. The next step for DA is peerdas, and the people working on this are almost completely disconnected from the people working on SSF or EVM implementation improvements.
I do think that'rollup-centric roadmap is complete 'is premature to declare victory. We are only 1.3x closer to blob full data (= 21600 blob/day) usage spikes. We need to start peerdas and start scaling security while L1 improves. "
Hoffman continued: "If we increase the block times for L1, don't we also increase the supply of block space? Focusing on L1 feels synergistic with focusing on L1 and L2 at the same time, and focusing on DA is only good for L2."
Vitalik explains this: "If we reduce the slot time without changing other variables, then the answer is yes. But the limit of both is determined by the bandwidth limit, so if we are satisfied with today's gas/blob limit, then we will also be directly satisfied with the gas/blob limit.
So I generally think of bandwidth and latency as two separate issues. I do think reducing L1 slot time is worth discussing, although it needs to be done carefully to avoid'killing 'independent stakers, or even geographically dispersed stakers. "