Recently, people keep asking me why on-chain data always "hiccups" for a moment... To be honest, it's mostly not the chain itself that's slow, but the data pipeline layer you're viewing. For example, Subgraph/indexers need to first fetch on-chain logs and then run mappings. When there's a reorganization or a surge of events, the index height can fall behind by hundreds of blocks, making it look like it suddenly froze on your interface. Another issue is RPC rate limiting, especially on public nodes. When traffic spikes, they give you a 429 or queue you up. The frontend usually caches the last result, so it looks like "the transaction was completed but the label hasn't changed yet." Recently, that wave of complaints about tool lag or misleading info, I think it's related to this: once the data source is a bit slow, the labeling system is very easy to be misled. Anyway, I now get into the habit of checking two or three sources, and if necessary, calling logs myself to confirm, so I don't get led astray by the dashboard.