I completely agree with what Guardian is saying, there are some issues with using the average improvement for the top 100 1500m athletes to predict how it would improve the performance of a 3:28 runner. That is why I have stayed away from trying to definitively predict how the spikes will effect performance on an individual or small group (runners under 3:29 for example) level. I simply stated that we have seen about a 1.5 second improvement since the introduction of super spikes in the men's 1500 and that seems reasonable to me based on the research and as a fan of the sport.
The benefit of looking at large sample sizes (100+) of race performances year to year is that we can account for some of the individual variation in response to super spikes and shoes. We have seen in multiple studies (including the one this thread is based on) that the individual response to super spikes and shoes is incredibly variable. We found that everyone improved their running economy while wearing super spikes, but the range of improvement was 0.5 to 4%.
At this point, all I can say is that the race time trends after the introduction of super spikes indicate that as the race speed increases, performance improvements decrease. To put a specific number on this would be overreaching in my opinion. We will unfortunately never have a study where the top 20 1500 m runners complete multiple time trials wearing the old and new spikes to have a definitive answer for how much the super spikes improve performance. However, I think we can use observational studies of race time and lab based research to make decent predictions.