A couple of weeks ago, Carson Fire of Elf Life started a new webcomic, Oopsie-Doodle, pretty much making it up as he went along and posting whatever he'd come up with that day on the website for people to look at. I didn't like it much, and just prior to the official launch, I posted a critique, telling him what I thought was wrong with the concept and the execution and predicting that it wouldn't have legs. That same day, he launched it formally and ever since, it's been building up a readership, which has been pretty good about donating money to Carson so he could make more (Carson has had severe financial troubles for years, as mentioned here before). It's not Penny Arcade and probably never will be but by any reasonable criteria, I was wrong and his new comic does have legs.
I still don't like it much, but that's neither here nor there. It's just not for me. What I do like is that rapid-development approach Carson has taken: make a few strips, post them, build simple website, receive feedback, make some more, refine the concept but leave the earlier, flawed batch up, collect more feedback, refine more, build out website. It's not how you're supposed to do it, but it's an approach that has some advantages. The immediate feedback means that flaws are corrected quickly and if the concept doesn't have legs, you can just scrap it after the first few updates, which will still have entertained at least part of the audience.
I have a number of non-ROCR ideas floating around in my head, and most of these would be easier for me to do than ROCR itself. ROCR, after nearly 20 years, is a comic with a lot of baggage and complexity - probably even more so for me than for a new reader who is faved with an archive of 1000+ comics. So the next few months, while I'm working on some pretty important changes in my life, would probably be a better time for trying out some of those concepts on that rapid-development model than for continuing doggedly with ROCR.