I think it depends on the task. Compare this to the difference between my wife and I carrying 100 pound sand bag to the end of the driveway and back. It's the same task but I will get it done quicker and probably without breaking a sweat. She will take a little longer and probably be sweating quite a bit. I believe for some tasks the higher end one will run cooler. Just my thought on the matter.
This is getting a bit off track from the OP's question, but I had to reply to say that's there's actually some truth to your logic on this.
If we look at Anand's results from the i7 vs i5 2013 MBA we see that at low loads the faster CPU actually produces better battery life, which means it uses less power during a given period of time. This should translate into lower heat output, as heat is a waste product of the energy usage by the system and should be proportional to power used.
This is most likely because it's a higher-binned chip that can maintain the same frequencies at slightly lower voltage. I'm going really off the track now, but if you could configure the frequency range of the faster chip to match the slower one you should see consistent efficiency gains at all work loads because of the lower voltages. Not easily (or at all?) possible on a rMBP, but still kind of interesting.
Unfortunately in real world terms this is largely academic since we're seeing a 2% increase in efficiency (and again only at "low" usage levels). I doubt anyone could really detect a 2% drop in temperature without measuring it.
Here are the stats I referenced for anyone who hasn't seen them:
http://www.anandtech.com/show/7113/2013-macbook-air-core-i5-4250u-vs-core-i7-4650u/4
In practice the faster system's exterior did get warmer (though only slightly). It can hit significantly higher peak temperatures, but tends to mostly even out over a long period of usage.
As a direct answer to the OP's question: No, in fact on average the faster system will likely run just a bit warmer.