[jschwartzi]

Does the number of mutations in a genome increase with the size of the genome, or with the length of time the genome is "in use?" Let's assume that the mutations are evenly distributed throughout the genome. If genetic mutation count is dependent on the size of the genome then it fits your "unreliable transmission" model.

However, if genetic mutation count is time-dependent but totally independent of the size of the genome, then having a larger genome actually does protect you from individual mutations, and it would do so exactly using the mechanisms described previously.

Think of two genomes, one large and one small, both existing throughout time. Both will accumulate a similar quantity of mutations from mutagenic processes which are time-dependent like radiation exposure.