I find using ticks to be more comfortable than a float to measure time, but is there something about using Time.get_ticks_msec() that is worse than just delta? Like as an example precision, is delta more precise than ticks?
Yes, but not it the way you may think. Ticks are real time and delta is your simulation time. The engine may not align processing function calls with real time. They can be slightly out of sync. From the simulation perspective, real time will look “incorrect”. So using delta is usually a better choice in processing functions. That’s why the engine provides it to you. This discrepancy will typically be more pronounced in _physics_process(), depending on your physics tick rate, while _process() will run closer to real time.
If you’re interested in more details, there was a discussion about using real time in _physics_process() some time ago. I contributed this in-depth explanation