Wouldn't it be the opposite? When Safari & Chrome were more similar you could get away with testing only one. Now you have to test on at least both Safari & Chrome, and anything that works on both of those is more likely to work on Firefox than something that only works on Chrome.