Mike, I see nothing wrong with your calculation either, and I think there are two aspects to this.
I suspect part of the problem is in the interpretation of the 'average bike ride' which leads to the figure of 300 calories per hour sometimes quoted.
The kind of average bike ride being envisaged probably wouldn't be on a windless day on a perfectly flat road, but would likely involve a light breeze and some undulations. Once you start climbing any sort of hill, the energy consumption shoots up.
If we model the hour bike ride as something like 45 minutes of it spent cycling up gentle 5% slopes at 10kph and 15 minutes rolling back down at 30kph, that gives the same average of 15kph. With a light breeze of 6kph and the 'average' less than ideal tyres/surface (coeff. of resistance set at 0.01), the calculator now gives 165 watts for the hour rather than the original model's 36. That equates to 142 additional calories required, rather than 31.
But these figures would be true of a 100% efficient machine, which a human isn't of course. For the average Joe to supply 165 watts at the pedals, he would most probably be burning up an awful lot more internally. Much is given off as excess heat from the body, as all we cyclists know only too well! The average human is perhaps only 25% mechanically efficient, if that?
It's then easier to see how 31 calories on an ideal flat road with a 100% efficient machine might translate into ten times that on an undulating road powered by a human engine.
We know that 31 calories would be nowhere near enough, because a 70kg man burns up 60 calories per hour when simply sleeping -- something I'm about to put to the test incidentally.....
