An RFDR setup that might work better:
It should be an entirely separate forum. Maybe not divided by genre, but then again, enough RFDR requests go through the Fantasy/SF/Horror queue alone to make it work, so perhaps genre isn't a bad way of sorting.
Every author who wishes to be critiqued submits the novel's first few chapters, a blurb (less than 500 words), wordcount and deadline, and a brief description of what sort of critique they're after (developmental edit, copy edit, focus on a certain subplot, holes in the premise or worldbuilding, basic grammar [sigh], style polishing, etc.). These are accessible not as a weekly list, but for, say, two or three or six months. Or everything that’s still open, oldest at the top. Or whatever makes a reasonably sized pile.
Every critter who wishes to critique must send crits of four of these opener chunks. These four crits are archived by the critter's name, so people can see a sample of their critique style before offering them an entire novel. Any chunks the would-be critter has submitted as an author himself are also archived there. Not everyone is compatible, and it would save a lot of frustration to know that before major amounts of work have been done.
No one can submit an RFDR of their own until they have completed at least one. (Authors need to understand how much work they're asking for!) Then, each submission has slots that can be filled by critters signing on, with the author's approval, UP TO as many RFDRs as the author herself has completed or signed on to do concurrently! When these slots are full, the opener chunk, blurb, etc., drops out of the manuscript pool and people work away busily, exchanging their emails outside of the hosted system.
Once a month, authors are reminded to award partial credits for however much their critters have done, and they notify the system whenever someone completes a manuscript. Authors may bump critters off their work at any time, although I don't think that should re-open the slot. [In theory, we had our chance to vet people beforehand, so we shouldn't gain anything from getting rid of them. This is mostly included (1) for people doing direct reciprocation, who realize they are wasting their time with each other and want to get out of critting the other guy's work, too, and (2) for critters who inexplicably start attacking an author's person. Since this is the Internet, it happens.] Possibly grumpy authors who bump a large percent of their critters ought also be flagged.
A count of partial and complete crits goes next to the critter's name (along with their 4 samples and any opener chunks of their own, including past and closed ones -- the point is to see how they write). If partial, there's also a flag saying whether or not they're still working on that piece, or whether they've been bumped or run past the deadline. That way, if someone is too flighty to ever finish what they start, they'll get some credit, but their tendencies will show for authors considering them in the future. Partial crits can be used to raise the crit count in the short-story section (the main part of Critters as it is presently structured), but they don't count towards an author's future available slots on their own work, or perhaps it should take several of them to count.
Beyond the slot system, I strongly think there should NOT be a "weekly crit ratio" requirement for RFDR. It's a hardship and a distraction for either critiquing or creating longer works. If I've spent six months working on three other authors’ novels and then taken off for a year to write my own, I want 3 slots, plus a couple more for works I'm critting now! If I'm constantly having to stop and fire off 201 words of half-hearted comma advice in scattershot critique of little stuff, I can't concentrate.
I think this would work a lot better than the present scabbed-on system. It would save a novel's strongest hook from getting buried in the weekly avalanche of slush. It would leave people the freedom to structure their periods of creating and critiquing more organically without "losing credits." It would prevent freeloaders asking for RFDR when they don't respect the amount of commitment required, and hopefully it would cut back on the number of dilettantes agreeing to RFDR when they have no gumption for follow-through. It would make it easier to judge compatibility up front. It would tend to produce direct reciprocation, which gives better results, as a relationship is formed.
In My Opinion. Of course.