There are different physical needs for current and voltage. Trace width depends on current. Trace spacing depends on voltage. I guess if you have multiple high current traces you'd want to space them more.
There are two other things to consider when designing a trace width based on current. Copper weight and temperature rise. The heavier (thicker) the copper, the more current a given trace can hold, and the higher the temperature rise allowed, the more. Most designers design for a 10°C temperature rise, but charts go up to 30°C, as in an ambient surrounding, this is ok (board not in an oven).
For instance, take a .250" wide trace specification (per mil std 275).
10°C rise ---> 5.0 A for 1/2 oz copper, 8.3A for 1 oz copper, and 12.3A for 2 oz copper
20°C rise ---> 7.2A for 1/2 oz copper, 12.3A for 1 oz copper, and 20A for 2 oz copper
30°C rise ---> 9.0A for 1/2 oz copper, 15.0A for 1 oz copper, and 24.5A for 2 oz copper
Note: 10°C = 18°F, 20°C = 36°F, and 30°C = 54°F (these are delta temps, not absolute). What this means is that if the ambient temperature is 77°F, then the copper on the board will be 131°F for a 30°C rise).
These are assuming external traces... internal traces (on a multilayer board) would be reduced current.
I did a 70A battery charger, and used 4 oz copper for the main board. Mind you, this was pushing 250A peak, duty cycle adjusted for an rms of 70A. You can average your current over time, if your trace width isn't so thin it acts like a fuse during the spikes.
Now, are these absolute limits? No... who says you can't use 30A through a .100 inch trace (100 mil)? Your system just has to be able to handle the heat.