Have you figured out the energy cost vs energy gain?
Do this by estimates or tests.
Perhaps buy some photo diodes with heat-shrink over them to make them more directional to simulate efficiency of PV beamwidth.
Collect some data and make a spreadsheet of energy gain for your weather patterns ( sun, cloud time of day)
-estimate Power lost by not having ideal tracker
- compute energy consumed to move x degrees.
Estimate wind force on PV panel at twister wind speeds.
Calculate torque on tracker arm and PV mechanical mounts, and estimate friction resistance and estimate total energy.
It than you expect to track continuously.
Then come up with efficient algorithm for tracker motor.
A simple servo with multiple photo diodes and comparators may be trivial, but a net gain in energy may not be trivial.