It is done by comparing the cohort who did the question with their results on the Core compared to other options. Germany, Speer, Leni, Conflict in Europe (which is the most popular now - just overtaking Conflict in Indo-China a year or two back - but very much similar) have the full range or students so no internal moderation necessary.
It is with the very small candidature topics e.g. the Japanese personalities often have fewer than 100 candidates between them. I have known years with no one doing Yamamoto and only a handful doing Hirohito and about 20 doing Kita so these are then compared to their other topics to ensure that they are internally weighted fairly.
The standard is not higher just because there are more people doing the option - the standard is marked the same regardless - often with the same marking criteria and same markers e.g. I have marked Speer and Leni and then moved onto the Asian personalities and marked all of the aforementioned Asians and Ho Chi Minh, Arafat, Jinnah etc. All marked to the same standard and on the same criteria but ... maybe the question suited one personality better than another so that personality's cohort may have lower marks compared to what they may have gained had they done a different personality.