I just read a fascinating article about how people's political views change how they actually perceive the temperature. I have to say that even though I know the mean temperature of the earth is very slowly rising (due to empirical data gathered by scientists), I perceive that the temperature is rising very quickly. And it turns out this may be due to my political beliefs. Always good to check myself before I wreck myself.
It may also be how temperature is perceived. While you're right the "average" temperature is rising very slowly, we do not perceive the "average". We perceive things like "holy crap it's still 80 at midnight" and "wtf? its 95 AGAIN?" Those things are entirely different. Your perception of summer heat is quite real.
If you look, for example, at the july 2012 weather summary, which can be pulled from http://www.nws.noaa.gov/climate/getclimate.php?wfo=iln, you can see that the average mean temperature (81.2) was 5.3 degrees above normal (75.9). That's warm. But we do not perceive mean temperature. The average MAXIMUM temperature (93.1) was fully 7.5 degrees above normal (85.6). That takes it from the average afternoon working outside thinking "pretty damned warm but its July" up to thinking "crap its #$!@ing hot out here". In practice a 7 degree difference between, say 79 and 86 is not nearly as noticeable as when it actually goes well over some individual threshold temperature that is clearly perceived as "too hot". For me, that threshold is around 91-92, for someone raised in, for example, Coastal South Carolina, that threshold may be higher, but most of at the start of July were acclimatized "Ohio in a mild June" and that threshold was NOT as high as it probably is now. The perception of heat really is enhanced above that rough threshold temperature. And also, that's an average max, which means that much of the time it was even hotter (up to 104!), and even on days when it was cooler than the average max it was possibly still hotter than the "normal" and still possibly hotter than our acclimatized threshold.
Also if you look at the average daily minimum we were at 69.4 which is only 3 degrees over the normal average of 66.1 and usually that happens when we're asleep. But, consider, in July the dewpoints here are typically in the mid-upper 60s and often higher. And, the minimum temperature is strongly related to the dewpoint, the warmer nights are also more humid, so you really feel the difference more. What that means to perception is that if you (assuming no AC) open your window at night, and the LOW temperature (usually reached at 5-6 am) is 70 and the dewpoint is typical, the rate at which your house cools off to sleep FEELS quite a bit different than on nights it cools to 66 or lower. It feels like "its just NOT COOLING OFF TONIGHT DAMMIT AND ITS STICKY AS HELL, THIS SUCKS" instead of "Its kinda warm tonight, better turn the fan on".
And, of course, that number is taken at the airport. At my weather station (carefully placed in the shade on the north side of the house) in the middle of the urban heat island, the daily high is always 1-2deg higher and the nightly low is always 3-5 deg higher (unless there is a good wind that day/night). Which makes it feel even worse, especially at night.
But, I think that what is really most useful in the context of perception in this data are the cooling degree days totals. This is (VERY roughly summarized) a measure that is designed to give some idea of how much cooling was needed to stay comfortable. It's a handy way of telling how much cooling people need to compensate for the additional heat in their homes. And its calculated by a running tally day by day and NOT an average or a point maximum, which makes it perhaps more useful to tell how hot a day actually feels. A day it stays hotter longer but has a lower maximum will still show up as more CDD. I do not think it's useful as an actual number, but it IS useful when compared to previous years monthly averages and to the 'normal' value. It's also used by the energy companies to usefully estimate how much energy people will/have use/used on a given day or over a month because there has historically been a very well established correlation between CDD and electricity usage.
The july 2012 CDD total was 510, which is fully 173 over the 'normal' of 337. That's a difference of fully 51% over normal. So by that admittedly perhaps not scientifically rigorous measure, people used their AC systems quite a bit more than usual for the 'normal' july. That's pretty huge. Last year was 503, which is also pretty huge. In other words... it was compensating for the uncomfortableness and NOT just for a small difference in average temperature.
It's also worth noting that for these NWS summaries that the "normal" was revised upwards last year to the averages from 1981 to 2010. If you use the "normals" for the 1971-2000 data you'll find our observed values for last month have an even greater departure from that "normal", which, admittedly might not make that much difference in practical memory. I don't think any of us really remember well what it felt like 30 years ago even if we THINK we do remember it.
So, I think it's actually re-written memory that is what is being measured in that study from the ars technica article and NOT at all the perception of current conditions. But that begs the question of how the memory is being re-written. Are some people mis-remembering it being cooler in the past or are some people mis-remembering the past as being warmer than it was?
But the truth is, while we may not really remember well how things felt in the past, and its really difficult to remember comparison of "feelings" from year to year, the perception of the weather being GENERALLY unusually hot THIS July was quite real in a very concrete and measurable way. The numbers do not lie. It was a damned uncomfortably hot July by this Ohioan's educated perception.