Though I wasn't able to find it offhand, the study with the inverse U curve showed that the moderate training group had lower all cause mortality than the extreme training group (who still had lower all cause mortality than the sedentary population). The meta-analyses I was able to find did not show the inverted U, but it is very possible that too few studies had "extreme" training groups for the far side of the distribution to show up.

If you only watch TV or read while walking at an incline on the treadmill, it's easy to rack up the miles without inconveniencing yourself. I probably hit about 24 miles a week doing this.