In my hands, the calculator has gotten me close but I had to develop my own sort of balancing refinement.
Mostly, I'm looking for low concentration components. I analyze alcoholic beverages so if a little ethanol bleeds through the switch, I don't care. My systems are 30 m x 0.32 mm DB-5-type precolumn going into the switch. I have a 30 m x 0.32 mm Wax-column coming out of the switch and going to a MS detector. The transfer line to the monitor detector (FID) is as I recall 2 m of 0.18 mm DB-5 type fused silica. Maybe it's only 0.10 mm tubing. I'm remembering this so bear with me here.
I choose a sample that has some analytes that are big enough that I still want to get 100% transfer to the second column. I then use the calculator settings as a starting point. I run that sample in heartcut mode (100% diversion of the chromatogram to the second column) and see what happens. I then change the head pressure on the precolumn (leaving the midpoint pressure alone) until I get the desired recovery of the peak of interest and only minimal bleed of the really big components (like ethanol) to the precolumn detector.
If I can get 100% of the peak of interest, it's been my experience that I get 100% recovery of everything smaller than that. It's a bit of a balancing act between optimum flows and good peak shapes - especially for the early eluters.
Here's an example of the kind of chromatography I get with my systems (I have several). It's about the 3rd post down with the graphics:
https://www.chromforum.org/viewtopic.ph ... or#p380164
Separate coeluting (on the DB-5) menthol from camphor.
I believe the system I describe above has 9.5 psig as the headpressure of the first column and 4.3 psig as the midpoint pressure (feeds the Wax going to the mass spec). I run in constant pressure mode. I've never tried constant flow with the switch. I get what I need with constant pressure.