I previously owned the Kleos and still have the Atlas and the Titan i. I am no expert but I hope to share my experience. If your results differ, we can learn together. There are many different strategies amoung audiogoners so try to see what get you the best sound.
From my experience with the Kleos using various arms (unipivot and gimbal bearing), some degree of antiskate is needed to bring out the best in the Kleos. The optimal VTF setting is within the factory recommend VTF using some antiskate. If I set VTF without any antiskate, I would arrive at slightly heavier VTF to counter the problems with missing tracking. This would result in suboptimal alignment of the coil and reduce the maximal performance achievable. The primary purpose of proper VTF is to alignment the coil precisely. Remember that changing the VTA setting will change the VTF and changing the VTF will modify the SRA with the same VTA setting. Proper SRA reduce distortion and VTA is the dial that you use to adjust SRA. In a sense, it is like solving a equation with two varibles that are interdependent. A good way to start is keep VTF constant, say 1.8g at record level, then change VTA up and down to find the proper SRA. As you are changing VTA, you will need to confirm VTF and may need to adjust counterweight to keep VTF constant at 1.8g. After that, repeat the same exercise with VTF at 1.75g, 1.7g etc. see what setting you ultimately prefer. This would land you a VTF that is slightly higher than optimal if antiskate is not compensated for.
Next set antiskate so both channels are tracking equally well. If listening to orchestra, the violon section should be playing as well as the cello/heavy strings section. If too little antiskate, the heavier strings lose focus and you cannot hear the sublties in play style, the boundary interactions within the acoustic space. If too much antiskate, the damage would be done to the violin sections. If you listen to vocals, the voice would be pulled left if too little antiskate and pulled to the right if too much antiskate. The centering of the voice is also affected by proper azimuth setting which is a potential pitfall. Best to listen for other clues like boundary interactions within the recording. In vocals, if the singer is centered, the boundary interaction of her voice should be apparent on both sides. If antiksate is off by large amount, tracking distortion would obscure the spatial clues present in the recording in the particular channel. Another way is to use a very difficult to track musical passage and having mistracking happen at both channels at the same time.
My understanding of skating force is limited but I know that it varies along the record surface. Tends to be least on the outside and most at the inside. However, it does not increase linearly towards center. It is modified by friction betw stylus and groove ( which is dependent upon speed of the rotation, shape of contact point of the stylus, VTF) and tracking error. Someone worked out a formula at vinyl engine if you are interested. A happy compromise is key here.
With the antiskate set, repeat the exercise with the VTF and VTA. Hopefully this yield the best sound.
Using fozgometer is OK but I feel this is a rather expensive device for its limited function. A PAA3 spectral analyzer can measure crosstalk more precisely and have many other useful functions. The lyra carts generally have great quality control, probably best I have had. My Kleos had better than 30db separation in both channels. The maximal channel separation you can achieve can be impaired by poor alignment. In the unipivots that I have tried: graham and centroid; the antiskate also modified the optimal azimuth setting slightly. Proper antiskate maximize channel separation.
As you alread know, the actual VTF tends to be lower than the VTF measured with the meter on the platter. I put a record on the platter and put the meter on a stack of cd/cd cases so the meter is leveled with record surface. Do your measurement there for a closer approximation. ( The actual VTF will be varying across the record surface as the record increase in thickness towards the center).
There are so many compromises one need to make during setup. I am amazed that vinyl stills sound better than digital for me.
good luck
