I don't think rlist gives you any benefits here.
I simplified your map function to return a simple vector of the variances, along with the simple vector of your image names, thats all you need to make a trivial dataframe/tibble to access in the normal ways.
image_variance <- images %>% map_dbl(function(x)
{
im <- load.image(x)
im %>%
resize(round(width(.)/10), round(height(.)/10)) %>%
imsub(y<187) %>%
as.data.frame(.) %>%
mutate(channel=factor(cc,labels=c('R','G','B'))) %>%
select(x,y, channel, value) %>%
spread(channel, value) %>%
mutate(new = R*0.2126+G*0.7152+B*0.0722) %>%
summarise(var = var(new)) %>% pull(var)
})
hist(image_variance)
(images_info <- tibble(images=images,
image_variance=image_variance))
(snow_coverimgs <- filter(images_info, image_variance > 0.008))