Aprostadyl theoretically works by relaxing smooth muscle in the penis and dilating the arteries. But that does not mean that that is a natural way of obtaining an erection ( I don’t know the role, if any, of Prostaglandin E1 on normal erection physiology). In a kind of absurd analogy, just because Viagra causes an erection does not mean that ED is caused by a viagra deficiency. (Or even by an NO2 deficiency).
I think you are oversimplifying, because if your theory were true, the strongest effect of NSAIDS on erection would be when they were in high concentration in the blood. But even high doses of NSAIDS do not have ED listed as a possible side effect.
The study talks about a chronic effect from NSAID use, not an acute ED problem, and I think that people who use NSAIDS regularly are in general, unhealthy and in frequent pain, and that would explain a lot of the ED. Even the study admitted that the reasons for taking NSAIDS is a confounding factor in the results.
Also consider that 93 cases per 1000 patient years means that most of the NSAID users do not have an ED problem, and we know nothing about the people who did have a problem. Were they taking massive doses of VIOXX for many years? etc. Again, this study does not justify a concern over taking aspirin or ibuprofen on an occasional basis.
For another absurd analogy, Im sure you could do a study that shows that cheese eaters have more ED than non cheese eaters, but the cheese eaters with ED would likely be the ones with the highest amount of cheese in their diets (causing cardiovascular disease), but that would not mean that if you eat cheese you get ED.