I noticed this yesterday on a brew. I had a wort with about 25% wheat malt. The crush I get with barley normally gets me somewhere around 75-78% efficiency. But with the wheat in mix, it fell to about 70% which is enough to get me to notice.
What was happening was that the barley was crushed normally. The wheat on the other hand is a touch smaller and since it has no husk to catch on the mill, using a Corona type mill, some of it was falling through or not getting crushed nearly enough. Plus, being harder, I think it forced the plates apart just a little bit more than the barley which contributed to the poor crush.
In the future, I'll do one of the following: crush the barley first at the standard setting, then crush the wheat at a tighter setting, or double mill the wheat and do the wheat first. It looks like if the wheat and barley are mixed together, the wheat malt pushes the plates farther apart which causes some barley to fall through the mill poorly crushed.
The other option is to lower the extract % to account for the decreased conversion on the wheat and just use more of it as a result.
It sounds like this is an issue with both roller mills and corona type plate mills as well. But the design of the corona mill seems to make this even more of an issue due to the inherent wiggle in the low cost clone.