If the volume of citable items during a two year period increases at a fast enough rate (instead of remaining fairly constant as it normally does in a regular journal that publishes periodic issues with a fairly set number of articles per issue per year), then the impact factor is likely to decrease simply because there is a greater dispersion of articles to cite. This is further complicated by the growing number of citing documents in the year after the two year period, especially if the average length of reference lists remain fairly constant per article and do not match the growth rate of the other two items. The IF drop then seems to be a simple inflationary problem.
If the volume of citable items during a two year period increases at a fast enough rate (instead of remaining fairly constant as it normally does in a regular journal that publish periodic issues with a fairly set number of articles per issue per year), then the impact factor is likely to decrease simply because there is a greater dispersion of articles to cite. This is further complicated by the growing number of citing documents in the year after the two year period, especially if the average length of reference lists remain fairly constant per article and do not match the growth rate of the other two items. The IF drop then seems to be a simple inflationary problem.