So of course now that we understand that the cases are not tested and just the material--it makes sense! We measured power density levels all around the case-the shielding material most likely isn't used "all over" because then the phone couldn't receive signal and wouldn't be able to engage in a call.  That's why we did not see even close to a 99% reduction when some cases were on the phone.  In fact, watch the video and you'll see some readings are more than 20% higher with a case on vs the naked phone.

The guidelines recommend keeping phones away from the body when they’re not in use—in a backpack, for example, rather than a pocket—and sleeping with phones away from the bed. People may also choose to use speakerphone or a headset to make calls, rather than holding the phone to their heads. (They should remove their headsets when they’re not in use, though, as these devices also emit small amounts of RF frequency.)
Instead, we have to rely on “observational” data, tracking people’s real-world cellphone use and their disease incidence. Studies using observational data tend to be weaker, messier, and less clear-cut than experimental studies like RCTs. They can only tell us about associations between phenomena, not whether one thing caused another to happen. So that opens up a lot of the ambiguity we’re going to delve into next.
Exposure to ionizing radiation, such as from x-rays, is known to increase the risk of cancer. However, although many studies have examined the potential health effects of non-ionizing radiation from radar, microwave ovens, cell phones, and other sources, there is currently no consistent evidence that non-ionizing radiation increases cancer risk in humans (2).