Add FAQ - sizing warning for single feature experiments

[eliserichards]

Related to mozilla/experimenter#9641

Add FAQ or explanation (maybe in data sections?) about how to adjust sizing for single-feature experiments that are targeting a feature that is already enrolled in Live multifeature experiments:

if data scientist run normal sizing for feature in “url bar” and get x% as desired typically, but if y% meeting the targeting criterion has been taken by active multiple-feature experiment where ‘url bar’ is one of the feature, they should then use x%/(1-y%) instead?