> there's no non-exposed control group that epidemiologists can easily compare to.
Incorrect. Time of onset is a control.
Look at the results on breast cancer from when they stopped giving post-menopausal women hormone replacement therapy.
Different countries and areas of the country get different generations of cell phone and towers at different times.
While there may not be a huge amount of funding going into this, the result of finding an effect between cell phones and biology would be so huge, that researchers in relevant fields very much pay attention.
If someone they thought they could get any correlation, even with massive p-hacking, someone would publish it.