molecular genetic evidence for a multipurpose genome

Just concentrate on your predictions. Even if there are many potentially lethal genes if knocked out, your prediction can hold good. Maybe we can prove this--lets say, get 100 zygotic mice and knock some 10% of their genome out and if you're right, then many will survive the experiment. To the lab!

Cool! Anyway, their studies demonstrate that up to 9,3 percent of the E.coli genome is irrelevant junk. How about knocking out introns? Or maybe using mice / Drosophila. These experiments might be off topic but they can answer other questions i.e. how much of the genome is junk, and how much are supposed to be junk/selfish but actually not.