Recently someone asked me: why does on-chain data sometimes "pause" briefly, even though the blocks are already out? Basically, it's not always the chain being slow, but the layer you're viewing is slow: the frontend first makes an RPC call to fetch events, and when RPC rate limits, it has to queue; the indexer/Subgraph also needs to fetch logs, replay them, write to the database, and then serve queries. If any step hiccups, you'll see prices/positions freeze for a moment. Layer 2 is now constantly debating TPS, fees, subsidies, but I think the stability of these "data pipelines" has the biggest impact on user experience. No matter how generous the subsidies are, users will still get annoyed when they see loading spinners. Staring at these things for a long time makes my eyes sore and my neck stiff... I now prefer to refresh a little less often, keep key data locally for verification, to avoid being "paused" and influenced by those brief hiccups.