<snip>
You've got a problem with that?
Not remotely correct. There are enough RNA-based mechanisms in our own cells to suggest that RNA preceded DNA, and RNA-genome viruses do seem to be hangovers from that RNA-world.
Where's the dilemma? DNA works by having the information it stores translated into shorter strings of RNA which then go off and turn that information into proteins or use it to regulate enzyme ativity. Quite how the original RNA machine acquired the capacity to translate itself into DNA and decode that DNA to make more reliable copies of the original DNA is an interesting question.
Why evolution didn't come up with error-detecting and -correcting codes as well (or instead) is an even more interesting question.
It saves time. There's nothing fertile in coming up with silly ideas that aren't likely to go anywhere. The scientific method does include a number of devices for throwing out bad ideas early so that only a few people get distracted by them. Peer-review before publication does exactly that. I haven't refereed al that many scientific papers, but most of them were un-publishable rubbish. I'm rather proud of the one that got published after the authors corrected the mistake that I'd objected to.