Speaking of sensationalists, Apple enthusiast John Gruber writes
As Alex Dedalus points out on Twitter, to say this is a crap headline is give crap headlines a bad name. Celsius and Fahrenheit are relative temperature scales, not absolute, so you can’t do percentage-based comparisons. Think about it: 33.6 / 28.3 gives you an “18.7 percent” increase, but if you do the math with the same temperatures in Fahrenheit, you get 92.5 / 82.9 = “11.6 percent” increase. If you really want to do a percentage based comparison, you need to convert to an absolute temperature scale like Kelvin, which shows you that it’s actually a 1.8 percent increase in temperature (306.75 / 301.45). This is middle school science.
There is some truth in this: the percentage comparison is quite crude, and potentially misleading. But there is also a fair bit of nonsense pseudoscience in the Gruber / Dedalus response.
Yes, Kelvin is an absolute scale, and Celsius and Fahrenheit are relative. However, all that means is that Kelvin is linked to kinetic energy regardless of the material. So its zero is linked to minimal entropy. Unless your average iPad user cares greatly about kinetic energy regardless of the material, the difference between absolute and relative scales is immaterial.
Why is the percentage smaller? Well, measuring temperature in Kelvin gives you bigger numbers. So, of course, when you do a crude division, the percentage is small. But choosing Kelvin as your measurement unit is as arbitrary as choosing Celsius or Fahrenheit. All of the percentage comparisons are equally meaningless.
To summarize: The Verge wrote a dumb article using a dumb comparison, and then John Gruber wrote an equally dumb reply with bad science. (Update: Jim Dalrymple falls for the nonsense too.)