I began The Jem Report as a site to hold my motherboard reviews so that I could more easily reference them when I was giving advice on forums. I wrote more than two dozen motherboard reviews over a period of two months. I developed my own performance testing criteria, tested every board, and did a writeup on it. In 2002 and 2003, these sorts of reviews were worth something because there could often be a drastic performance difference between northbridge chipsets and frontside bus speeds. The usefulness of motherboard reviews has diminished since then, however, and although I have motherboards here that I could be writing about, I find myself unable to come up with meaningful test criteria. AMD effectively killed the frontside bus and the need for a discrete memory controller when it introduced the Athlon 64. That means that there is no longer a reason to measure motherboard performance. What does that leave for the review? A list of specs? What could I tell you in a review that the motherboard manufacturer's site and the buyer ratings at Newegg.com can't tell you more effectively?
The answer is: nothing. I can tell you what the board has in terms of features -- I/O ports, number of PCI and PCIe slots, etc. -- but you can just as easily get all of that information from the manufacturer. Just go to the motherboard maker's site, navigate to the informational page on the board in question, and you'll find everything you need to know about the board in terms of technological capabilities.
But what about customer satisfaction? What if the board looks good on paper, but totally sucks once it's installed in a system? What if it fails after two weeks, or doesn't work with a certain brand or model of video card? In most instances you can go to Newegg and read through the customer comments and ratings and find all of these problems if they exist. These are from people who actually bought the board and have used it. Certainly not all of the comments will be useful, but there are usually enough of them that you can get a good impression of what the board is like.
Professional motherboard reviewers often get pre-production review samples from manufacturers; their reviews are based on these, not on the final release, which can vary significantly from the review sample. I know -- I've tested review samples in the past. Usually pre-production samples have more problems than advantages (such as reduced performance, RAM incompatibilities, and trouble with new connection standards or technologies), but it's not unheard of for a review sample to perform substantially better than the final release. There are also several revisions of every motherboard, each one introducing newer (cheaper) components, bug fixes, and expanded device compatibility. These different revisions almost always show a measurable (though rarely noticeable) difference in performance.
That brings me to my last point -- noticing the difference between motherboards of the same technological era. If you can't take the Pepsi challenge with two motherboards and notice a marked difference in performance, then there is no need for a review.
I've given it a lot of thought, and I can't find a way to make a motherboard review meaningful in this day and age. Therefore I will no longer write them -- not that I've been doing any lately anyway. If you want to prove me wrong, please do; the forum link is below.
Editorial commentary 
