Core 2000 experiment

I decided to do a little experiment with the core 2000.
I went trough part 2 till 10 in only 2 days. This wasn't very enjoyable of course and I doubt I learned much of the words.
But I interesting how the daily reviews are being affected by it.
The first few days were an overload of reviews of course. But now less then a week later it already starts looking a lot better.
It's still in the hundreds of course but I'm quite interested how quickly it will drop to about 100 a day.

I click easy on all sentences so they get pushed forward quite quickly. With the words I click hard on the ones I don't know and good on the ones I know.
This means the daily load will be mostly the hard cards in a few weeks.

I guess this would work terrible if I was actively trying to learn these cards instead of just clicking without really thinking. But I doubt learning the core words along the kanji works well anyway.

There a few ways to realistically deal with this :
1. Spend 100% of the effort on kanji first and then 100% on the words.
2.spend 90% on kanji and 10% on words, and after that 100 on the remaining words.
3. 50% kanji and 50% words and finish at the same time.

And the unrealistic way I'm dealing with it right now :
4. Spend my attention on learning the kanji at a manageable speed and spend time without attention on words in an unreasonable speed. And only start paying attention after I finish the kanji.

For the kanji 20 to 30 cards is quite doable. For the words I assume it would be an higher number as they also contain those same kanji. Let's assume between 50 and 100.

I will do some calculations later to see how long option 1 would take.
I assume option 2 and 3 will be slightly slower.
I'm not sure yet how I would calculate option 4 but I will think about it.