These days, someone asked me again, "Isn't on-chain data real-time?" To put it simply, a lot of the "on-chain" data you see is actually being relayed by the nodes/RPCs/indices you use. When the queue is backed up, caches are outdated, or indices haven't caught up, the balances/trades/events you see can be half a beat slow or even wrong by a beat. I personally have been burned when launching new protocols: even though the transaction was sent out, the frontend still showed no activity, and I accidentally sent another transaction... only to realize later it was an RPC hiccup.

Recently, before the upgrade/maintenance of that mainstream public chain, everyone was guessing whether projects would migrate. I think there's no need to rush into imagining a "mass capital exodus." Many judgments about whether to migrate are based on whether your data is fresh enough. Anyway, my current approach is pretty simple: run multiple RPCs to check for consistency, wait for block confirmations before making critical moves, and don't let "what I see on-chain" set the pace. Sigh, FOMO really is hard to control.