Now I think the case shows that not only the referee process is in dire shape, but that more broadly there is a problem in our scholarly culture. (It's not impossible that Stone case only reflects on a smaller field because folk in midieval, renaissance, and early modern do 'cover' a lot of territory. But I think the cultural problems that the Stone case illustrates are not confined to small area of philosophy.) Often the people that are in a good position to judge the quality of each other's work are not doing so. Moreover, and this may be hard to understand for those that work in US-style tenure, in many scholarly environments quality control (which will determine not merely publication, but also professional advancement, grants, etc) is now entirely farmed out to the often anonymous, time-pressed referees--moreover, the referee process is as good as its weakest link. (Recent research suggests that with even 10% bad referees the system cannot maintain quality control.)
As a not entirely amusing aside: it dawned upon me a while ago that if I ever wanted to be read by scholarly peers, I would have to start doing major "kissing down," that is, cultivate the attention of other people's PhD students! (I found that "kissing up" is a waste of time. Senior people are too narcisstic, overworked, and invested in their own projects to really be interested in one's competing programs--the few exceptions prove my point.)
So, in order to improve refereeing and our general academic culture, I offer two rather small changes to our current intellectual infrastructure. Both are offered in the spirit of properly alligning incentives (assuming that scholars are status-seekers) to prevent plagiarism and improve quality of 'output.'
1. All accepted articles should carry the name of the referees. (I believe this was first proposed by Alva Noe on the Leiter blog.) This gives them credit for the success and makes them more diligent to prevent fraud. Moreover, it also increases transparency about intellectual networks. 1b. Also publish referee reports. (With digital publishing this is getting ever more easy to finance.) In some cases these reports are quite elaborate or technical; why not make them part of an author's output (Maybe to be measured for advancement) Refereeing book manuscripts, especially, can be very time-consuming work. [Notice that this proposal still allows one to remain anonymous vis a vis rejected papers.]
2. A. As Kent Emery proposed (see link above) a serious book review should count as a normal article. (Both in tenure-granting environments and in European grant systems book reviews count for almost nothing.) A good book review should hold authors/editors responsible. They should also be interesting. (I spend a lot of time on reviewing -- although half dozen editors have every right to be angry with me about my current backlog [yeah, yeah, eventually I cannot hide behind recent switch of jobs and new fatherhood!] -- because I think it is useful and important work; I also believe book reviews are often more widely read than articles, see my point about "kissing down" above.) B. Journals should commission 20% of their 'space' to reviews of recent literature. (Again with digital publishing this should be cheap.)
Obviously both proposals will make us more focused on each other and less on the truth or the world. Since I am a big believer of the do no harm principle, this is probably a good thing.