Yeah something is definitely a bit borked. Wonder if there is more traffic in the federation or more users on the instance itself? Or is this just one of the joys of using a prerelease service? Lemmy is pretty resource-intensive because there's a lot of frankly pretty questionable design choices all over the project, from database design to coding practices, and it's apparently turned out to be a bit painful to scale
edit: none of the Lemmy stats I found looked spiky at all around this time, see eg https://lemmy.fediverse.observer/dailystats.
edit: sopuli's own stats look relatively flat too https://fedidb.org/network/instance/sopuli.xyz so the slowness is probably just lemmy-api getting cranky for some reason