In my opinion that Manske Step is not even needed. More severe it takes much more time than just settling Freebase Alkaloids for any full precipitation and then they are not easily filtered / separated.
I think the original idea was to get rid of Vasicin or Vasicinon - these carry an -OH or =O group and are therefore more water soluble - this will keep them in solution even at supersaturated situations like during Manske step.
They have a bad GHS Symbol on wikipedia, but actually that is only because they can induce premature birth at high level in pregnant women. Other than that there is absolutely no harm. And eating this stuff while being about to give birth might be a really scarce encounter
.
But on top I made an analysis and found absolutely no Vasicin / Vasicinon. So either (at least in my batch) there was not any at all or they are removed anyways in the way of not doing a Manske. That would be much easier and faster:
1. Boil your seeds at pH 3 (dont grind, Alkaloids are in the shell).
2. Dont reduce and add NaOH until reaching pH 11.
3. Wait for settling of Alkaloids and then decant as much water as possible.
4. Put in a big container and fill up with more water.
5. Repeat at Step 3. until the water is completely clear or pH reaches below 8. These 2 situations will most likely happen simultaneously, so color is also just a good indicator.
6. At last decant as much as possible and fill your residual Alkaloids on a plate, evaporate in oven at 80 °C with high ventilation
= pure Harmin / Harmalin mixture, no Vasicin / Vasicinon so Manske is unnessecary work IMO. Also not even a single filter step needed!
If you dont have any precipitate you could simply start at step 2 and still get your stuff back.
Other than that there cannot be too much of anything, you could simply add more salt if only "50 % of the water was saturated" let's say.