Lazy splitAt is too lazy?

[Lysxia]

There's a missing bang pattern which makes Text.Lazy.splitAt on the empty text lazy in the Int64 argument. In Core that causes a duplicate worker-wrapper for the first iteration which must be lazy, and the remaining ones which are strict.

Is that accidental? "laziness zeal"? Is that going to be a breaking change?

splitAtWord also has this problem but it's internal AFAICT.

splitAt:

splitAt :: Int64 -> Text -> (Text, Text)
splitAt = loop
  where
    loop :: Int64 -> Text -> (Text, Text)
    loop _ Empty      = (empty, empty)
    loop n t | n <= 0 = 

[Bodigrim]

Data.List.splitAt has notes about laziness, it is not lazy in n. I think we should follow the suit.