standard_error() for merMod with correlated random effects fails

[strengejacke]

library(lme4)
m <- lmer(Reaction ~ Days + (1 + Days || Subject), data = sleepstudy)
parameters::standard_error(m, effects = "random")

[strengejacke]

This also fails:

library(lme4)
sleepstudy$random <- rnorm(nrow(sleepstudy))
m <- lmer(Reaction ~ Days + (Days | Subject) + (0+random| Subject), sleepstudy)
parameters::standard_error(m, effects = "random")