Some papers say they study the identification and quantification of species involved in initial atmospheric corrosion. Consider the case of this article in Corrosion - The Journal of Science and Engineering, for example, it provides an exceptionally good model, but I am interested in the copper corrosion process as a whole, not just the beginning. What exactly is understood as "initial corrosion" and what's the relevance of studying such a short time-span? That paper's tests reach no longer than 300 minutes, while other studies measure time in weeks, not minutes.
Here's an attempt at an answer -
Initial corrosion may be taken to mean when the electrochemical reactions occurring are the main driving force for the material to corrode, meaning changing chemical and/or physical properties. As corrosion continues, physical factors (stresses, strains, grain defects, crystal defects, etc.) can contribute and dominate the reason why a material might fail. Studying the initial mechanism can lead to solutions that inhibit or prevent corrosion since it is currently not likely you can salvage a heavily corroded piece of metal and return it to its original service.
As for the duration variance of the studies, it could be that mechanisms change with time or that some systems are slow to corrode but still need characterization to allow for the material in question to be more accurately used. This, however, is more my opinion than actual experience.