Interesting article. I'm not sure, though, that it isn't a cyclical thing. Notice the ATP numbers just compared over the last decade; the average age in the 70s, for instance, was much higher - at least in the top 10. This is not to say that I disagree with the basic premise of the article, or what herios is saying - I basically agree - but that we also need to look at at least two further factors: one, generational cycles, and two, different levels of players.
In terms of generational cycles, as we know just looking back over the last 20-25 years, there are periods in which there are a lot of great players playing at or near their peak level, and periods in which few elite players are at that level. The late 80s to early 90s was a bit of an unsung Golden Era, in my opinion. People always talk about the late 70s-early 80s as the Golden Age--with Connors, Borg, McEnroe, and Lendl--but a decade later you had Lendl, Becker, Wilander, Edberg, and then Courier, Sampras, and Agassi all playing at or near peak levels. In mid-to-late 90s the field dwindled and tennis went through a bit of a "dark age"--at least in terms of top talent--until the emergence of "Fedal" in 2004-05, and then "Djokoray" in 2007-08.
The point being, talent comes in cycles, and in order to make meaningful statements about age trends we have to look beyond a relatively short period such as a decade.
In terms of the latter, I think we should look at at least three levels: all-time greats (multi-Slam winners), near-elite players (Slam challengers, top ten players), and top 100 players. Regardless of what is happening right now, as I've said before historically almost all all-time greats peaked around age 24-25 and gradually declined after that so that they become more like top 10 players in their early 30s. I'm not sure this has changed; the only current all-time great that we can really look at (that is old enough) is Federer, who certainly supports the idea of a peak around 24-25, a half-step back around 27, and a plateau until a further decline sometime in the 30s. We'll know more about current trends in the next year or two as Rafa, Novak, and Andy all reach that "half-step back" phase. Rafa is entering it now and COULD be showing signs, but it is hard to tell. I think we'll know more by the end of the season.
But this doesn't necessarily oppose the idea that MOST players are peaking a bit later. As I suggest before, it is a very light hypothesis of mine that great players reach their absolute peak in their mid-20s, while second tier players more frequently peak a bit later - or at least don't decline from their peak established in their mid-20s in the same way that true elite players do. This might seem counter-intuitive, but think about it: the difference between the all-time greats (say, Federer, Nadal, and Djokovic) vs. near-elites (say, Tsonga, Ferrer, and Berdych) is a combination of greater consistency and an ability to tap into a "supra-level" of talent, but the baseline talent level isn't that different. What might happen is that once these elite players reach their late 20s their ability to tap into that supra-level diminishes and becomes less frequent, so they start falling back towards "mortal" levels like those of the near-elites, whereas the near-elites never reach those supra-levels, beyond a stray tournament or two, so are able to maintain their best.
In a way you could say that elite players reach a high peak in their mid-20s and then come back to a plateau in their late 20s, sometimes into their early 30s, whereas near-elite players reach a plateau but never a higher peak and just maintain that level throughout their late 20s. This could be similar to the difference between someone who is brilliant and a genius; a genius is also brilliant--that's their "baseline talent level"--but they are more frequently able to tap into an even higher level of brilliance, aka genius, whereas the "merely brilliant" can also tap into genius, just not nearly as frequently.
If we look at a near-elite player like Tomas Berdych we see the promise of an elite level in 2005 when he was 20 years old and won Paris Masters, his only ATP 1000 or higher title. But instead of continuing upward, he plateau-ed for a few years, and then went up a level in 2010 at the age of 24-25. Just about every elite player, however, keeps rising in their early 20s.
Other near-elite players like Tsonga and Ferrer broke into the pro tour a bit later. Ferrer didn't reach the top 20 until 2005, the year he turned 23; for Tsonga it was 2008, also the year he turned 23. Ferrer has a few years outside the top 10, but both have been essentially been top 10 players since age 23, with Ferrer having his strongest run in the last year or two at age 30-31. Tsonga has pretty much played at the same level since 2008.
My tentative conclusion, or at least theory based upon the initial hypothesis, is that there's a crucial developmental range in the early 20s when it is determined whether a very talented player will become elite or not. If you haven't shown elite status by age 22-23, chances are you'll never get there; this doesn't mean that players can't improve after 22-23, but that they'll almost certainly never be elite unless they show that level in their early 20s.
A bit rambling, but I hope that makes sense! I might put it all together with further research in my much-neglected Tennis Frontier blog!