Not really. EPO testing is now possible; it wasn't in the 1990s. That's why we saw dramatic increases in performance on the track in the 1990s.
The real reason is more financial: see the graph of prize money referenced in the article, and also take a look at the current state of track 10,000m racing. More top talent is attracted to the marathon, away from the longer track races, than before.
It's true that record breaking has switch mostly from the track to the Marathon - probably because of financial reasons. But speeds haven't really slowed down (like they have in cycling).
Cycling has a biological passport which (at worst) limits how much a doper can improve their blood chemistry. Athletics doesn't have that, and doing point checks isn't as effective as looking at changes.
I don't deny that doping still goes on, but people are forced to resort to less effective means than they used to. Take the example of Rashid Ramzi--it is likely that he was not the only doper, just a more aggressive one. Why was he caught when others were not?
> speeds haven't really slowed down
This isn't wholly true. Have you seen a 3:26.00 1500m, a 7:20 3000m, or a sub-12:40 5000m lately?
We saw dramatic increases in performances on the track in the 90's because that's when the African training camps were formed. Instead of one Henry Rono suddenly there were 100. Though I wouldn't deny EPO use was rampant, times didn't slow down either once the EPO tests came out, so I think its effect was exaggerated.
The real reason is more financial: see the graph of prize money referenced in the article, and also take a look at the current state of track 10,000m racing. More top talent is attracted to the marathon, away from the longer track races, than before.