I happily defer to equill's data over hearsay. I've never directly tested the assertions from the LEM profile, so I have no direct experimental data to support or refute it. I've always just partitioned as if the assertions were true (couldn't hurt, might help).

I agree with equill that the symptoms are more consistent with some other problem. As usual, he sayeth much sooth.