My understanding is that these statistics show the coverage for each KEGG Module in the particular phylogenetic group. For example, M00001, is found in 275 of the eukaryotic organisms in KEGG (which represents 90.5% of the eukaryotes in KEGG). The same logic applies for other groups.
An quick example demonstrating this hypothesis, done with the Protists group found in KEGG Organisms (since it is easier to count by eye). There are 43 protists, and in the statistics it says (as of 2015/7/9) that module M00001 is found in 33 of them, hence: $33/43=0.7674419$ which agrees with the 76.7% shown.