I don't see where the canoe rack is described. You mentioned panel mounted on the roof rack, so panels are approximately at the same height as canoe rack, only soft cradles are few inches higher. Normally people use 2"-3" high supports under gunwales. If you keep 10-12" distance from cradles to panel, I don't see this causing much problem. 
Array size and batteries:
There is a rule 1W per 1AH, but it should really be calculated individually because a lot depends on your usage pattern. 2*6V batteries @225 AH each need 45A max. 4*6V need 90A max.  500W solar will generate 35A max. 
I think that you can have 4*6V with 500W because you will have an alternator in addition to solar. Also, if you conserve energy and use less than 85-90 AH a day with 4*6V, batteries will cycle in 80-100% SOC. At this point it only accepts low current and you will never need more than 20-25A.
Panel voltage is not the controller output voltage, or the battery voltage. This leads to discussion of 12V panels vs 24, PWM vs MPPT, and has been discussed here way too many times.
Single small panel has 18V, battery needs 13-14V. PWM controller understands only amps. It will cut the excess volts. These volts are lost. In other words, panel watts are lost.
MPPT controller transforms Watts into Amps W/V=I, where W is panel watts and V is battery voltage.  If you have 570W, MPPT will transform 570/14 = 39A after controller losses. Doesn't matter - what voltage, series or parallel, it will still convert 570/14=39.
MPPT conversion comes with losses 4%-10%, losses are higher when you feed higher voltage, but these losses are not terribly high. 
One problem with MPPT is that panel voltage drops in a hot weather. This doesn't matter for PWM because PWM cuts them off. But with MPPT these lower volts mean lower watts going into controller, so after conversion W/14=I the current will be lower.
You see now why people don't like discussing PWM vs MPPT. Many variables. The bottom line is - get more total wattage. The more, the better.