There are some very important points in this article by Richard Harris which relate directly to the state-of-the-art in EMF research:

Scientists point to what they call the “reproducibility crisis”—that is, studies whose results can’t be duplicated and are untrustworthy if not invalid.

and one that may surprise some people, but really shouldn't:

He showed that small sample sizes and bias in study design were chronic problems in the field and served to grossly overestimate positive results.

and:

Follow-up studies, they showed, overturned half of those initial positive results (though such disconfirmation rarely got follow-up news coverage).

This one may ring some bells:

Another key source of error is bad research design: Too many scientists conduct poorly conceived experiments or fail to analyze them properly. They often use too few animals and don’t take all the steps necessary to reduce the risk of bias.

But none of this should really be news: it has been discussed for some years with specific reference to EMF by Ken Foster and Joseph Skufca, amongst others.

so how bad is it? Estimates vary from 10% to "pretty much all of it", but a consservative estimate is probaly 25-30%. There are pretty good data to show that in general, poorer experimental technique and quality control tends to generate more false positives, which perhaps isn't a surprise.