>which has generally been refuted by linguists in recent years

The strong version has been refuted, the 'soft' version is pretty well accepted among linguists.