I tried once, late at night, scrolling through on-chain data to make a small story. Clearly the block had already been produced, but the page would seem to "pause" for a moment, then catch up after a few seconds. At the time, I thought it was just my internet connection acting up. Later, I followed the logs and found that there are only three common causes: RPC being rate-limited (requests queued or dropped if too frequent), the indexer hasn't finished processing the new block, or the subgraph is syncing/rebuilding, causing queries to hit old caches. In other words, the blockchain itself is real-time, but the data you see has to go through "transportation + processing + delivery," and any pause in between can cause perceived delays. Recently, everyone has been arguing about L2 TPS, fees, and subsidies, but I care more about whether these "data pipelines" are stable—because the most vulnerable part of the user experience is these uncertain pauses. Anyway, I now keep a backup RPC, and for key queries, I cross-check the block number to stay more calm.