Hi and welcome,
it is not uncommon that different methods have different factors for the same flow, in a seemingly similar impact category. For chlorine, you have it indeed in ReCipe midpoint but not in ILCD, e.g. in the category freshwater ecotoxicity.
We are about to release a new LCIA package btw. for openLCA quite soon, with updated and quite extended methods, which does not really solve your point though.