I don't know about commonly thought to be false because I'm not sure it's commonly thought about at all, but it's a theory with considerable weaknesses
Ekirch hypothesis: Early humans had two distinct phases of sleep with an important gap between. Lack of in-depth discussion or even a name for that gap in any language is actually evidence of it being so common it wasn't worth commenting on. It disappeared - again without comment - because of widely available artificial light, although it actually makes less sense to dedicate midnight hours to stuff like household chores and reading in the middle of the night without easy light sources, especially in northern European summer when it's about the only time there isn't daylight.
Null hypothesis: early humans slept much like today's humans, sometimes waking or being woken in the night, occasionally even intentionally but generally not making a big deal of it and trying to sleep through. People sometimes described periods of broken sleep as "first sleep", "second sleep" and even "third sleep" but commentary on the practise of biphasic sleep and importance of midnight waking is harder to find because most people didn't do it that way. A lot of references to "first sleep" can be found if you search digitized records with that string and its foreign language equivalents, but so can references to obviously non-systematic things like "first injury" or "first marriage". You can't generalise human behaviour in the absence of electricity from one tribe that does have a midnight break when numerous others studied don't.