Here is something more for Usi to consider, since I am getting sick of being lumped with lunatic Korean nationalists:
I actually think Japanese colonial occupation of Korea was fundamentally different in nature than Western colonialism, and in many respects beneficial; I think the "comfort women" issue is a dead one, since it was explicitly dealt with by President Park in the 1965 Normalization Treaty; I think the preponderance of historic evidence suggests--though I am tentative on this one--Dok do/Takeshima is not historically Korean territory; and I certainly do not think either the Japanese emperor or Prime Minister ought to apologize yet again for the colonialist experience.
Some kind of a Korean nationalist, eh--just like the original poster!

It's most ironic that I actually have to defend myself from charges of being a nationalist, given that I have to deflect accusations of being a Japan apologist among koreans
Of course. Right conclusions mean nothing if you reason badly.
And my own personal experiences are similar: In my early youth, I was programmed to hate and despise everything Japanese by the Korean educational system and group-think. Greater education, exposure--and I daresay "enlightenment"--have slowly transformed my view.
You are more reasonable. That's why I specifically named you in one of my post. I don't think I'm alone in that observation. One thing I tend to learn as I get more and more education is the skill of independent critical thinking. I've met some Asian graduate students where they don't question professor's teaching. I was shocked when one person responded to my question "why don't you guys ask questions in class" with "why? won't the professor get upset?" This is from someone who went to a prestigeous undergraduate school back in Asia. I explained that while I was in undergraduate, more than one professor, some in math/science, some in social science feel "depressed" when there are no questions in class. When students ask no question in class, it means either the professor's class is too easy, which is no good, or the professor is not clear or material too hard so the students don't understand, which is also no good.
South Korea has excellent grade school education, compare with USA, but I found some of my friends from east Asia tend to shy away from challenging the professors while in graduate school. Indian students are fine, they challenge professor. In higher learning, you probably should develop critical independent thinking. That's how new knowledge is made. If you blindly accept what predecessor said without challenge, we won't have new idea to prove and won't be able to develop and discover new theories.
I do partially buy your evidence this time, but I'm still unconviced.
At least I do doubt that there were exaggerations, just as in China the number of soldiers involved in battles tended to be exaggerated.
Or picture this situation: Nihonshoki said that Yamatai (Kyoto) had had a million people.
Would you buy that?
Check the source again.
There are estimates of Roman population at older times, and none is greater than a million.
I think it's hard to get accurate battlefield death/casualty number. For that to happen, all sides involved have to be truthful and there are no desertion (which happens fairly often until recently where it's much easier to keep tabs on your troops)
One of my roommate was a recon lance marine in the 1st Persian gulf war. When I asked him how many enemies he has killed, he told me he has no idea, then proceeded to explain why. They often recon at night to utilize our superior technology vis a vis Iraqi troops, using night vision goggles, etc. he was involved with some battles that made to newspaper headline but still, he has no idea how many enemies he has killed because when you are fighting in the frontline, survival mode often takes over. You duck behind cover, then just shoot from behind your cover. You don't poke your head out, check and count how many enemies are around you, how many you have wounded, how many have you killed, etc. It's too dangerous, you expose yourself to danger when you are ordered to, such as complete a task. Otherwise, it's duck, fire, supress enemy fire or kill them (directly or via support such as air bombing, strafing, etc.)
another one of my friend enlisted in the marine after 9/11. His first tour of duty was in Afghanistan, in the most remote front base we have close to Pakistan. He was decorated, he estimated he probably wounded/killed over 30 Al Qaeda but even then, he is not 100% sure. They're in hostile territory, you don't stop and count the enemy dead + wounded, especially they were ambushed often. How they got some of those number of him wounding enemies (he was a mortar man), was when it's safer, some air units, such as helicopter may fly by and count the body parts left behind by the enemy (mortar explosions can be messy). Sometimes you estimate, you have a rough idea how many enemies are around based on the enemy fire, if you silence them, they're probably dead or wounded.
I read some study on those ancient battle number. One way to estimate is by estimate how much food a particular nation/province/fief/duchy/etc can produce and then roughly translate that to how much population that "nation" can support, which in turn can give you a rough estimate of how much soldiers they can call on battle. You then tweak it by estimate how many are combat, how much are support. Mongolians are able to mobilize a lot of their population into their cavalry as their troops are able to live off the land better. Iranian heavy cavalry will require more support. bronze age infantry probably require less support so a nation should be able to mobilize more. I think US infantry used to require 4 support personel for every combat soldier on the field. That number is probably higher now.
Asia land combat # are probably bigger as rice is the best cereal so to speak. If I remember correctly, you can produce the most calory from 1 hectar of land by planting rice IN the southern part of Asia, where it's possible to harvest rice twice a year.