No, though of course it isn't that "simple".
And use of the term in the referenced article is yet another indicator of lack of rigour in it.
No, what many (probably most) people understand as "average" is the arithmetic mean. Other common averages in use are the Median, and the mode - either of which would be more useful in this context that the mean.
If the article had been based on any scientific survey or dataset, then the author would have specified mean or median, rather than just saying "average".
The mean is where you add up all the values and then divide by the number of value.
The median (probably the most useful in this context) is the value of the middle entry, so if there are 1001 values, you will take the 500th when they are placed in order.
The mode is the most common value among the entries - in this context, you would group speeds into bands of maybe 0.1 or 0.2 mph, and the band which most riders fall into would be the modal average.
Having said all of which - what I suspect
@fair weather cyclist was getting at above is that anecdotal evidence that most of us go way faster than his figures suggest means nothing in terms of disproving his charts. Partly because the people attracted to this forum are mainly going to be among the keener members of the cycling community, and therefore tend to be well above the median (or mean) for their age. And partly just because anecdotal evidence from a few dozen people really says nothing about the average (whatever type) of many thousands of people.
I don’t personally consider seeing a computer generated image of somewhere as actually going somewhere but you may beg to differ 
