This is an interesting topic. I think 20 copies of p53 itself creates another paradox. Since too much p53 would cause either apoptosis or too much senescence(depending on the allele and context). This would deprive the animal of dividing cells pool pretty quick, unless there are more adaptors for it and some other function. Such as in nuclear organization. But before going there I would rather ask that how many of these copies are actually expressed per cell.

Regarding the paradox in question , I think metabolic rate would make sense. But that won't explain the entire picture. I think rather than considering the entire lifespan one must consider the lifespan post reproductive age. Because even in evolutionary terms there won't be any selection against shit that happens beyond this point. Since life expectancy in elephants is 70 yrs approx and they can reproduce until age 65 or so this may explain the rarity of cancer in elephants. At least if you compare it to humans, lab rats, pet dogs all of which we make sure surpass far beyond reproductive age. In fact I would predict that any wild animal not going long beyond reproductive age cancer incidence should be lower. I think just comparing between wilds and correcting for life post reroductive age would be a cleaner picture.