It has been stated many, many times that specifics will be discussed on the drillbit forum. Yes, it is in a thread over there. Of course you're free to post here, but don't expect an answer.
Stuff you, I WANT people who bought the 8 board from this thread to know there is a USB problem, and to have access to the solution. It's common courtesy.
If you're referring to the libusb error; That's more of an issue with CGMiner, not the actual product, though there might be some firmware updates that could change the timing.
However, I can honestly say I've been running 6 boards for far over 72hrs added 2 boards to that setup 24hrs ago and not a single error, also, on a raspberryPi running MinePeon. However I did do somethings to achieve this, probably the major thing I did was change my difficulty at my pool as well as my queue on the miner software.
I figured the queue to be 4 per 20GH/s (so, 32 @ 8 boards) and I upped the difficulty to 64 at the pool. Since the difficulty is higher, the boards aren't vomiting information as fast (almost just as fast, but not as fast), therefore it allows for some small amount of breathing room between USB communications, where as it would normally just be flooding the miner software with messages.
Why did I do these things? Well, being completely honest, I hoped on my rPi for my Blue Fury miners and forgot the one thing I loved about BFGMiner versus CGMiner, it dynamically set the queue size and a few other variables, where as CGMiner, doesn't. So, I saw the setting for my 30GH/s on my Blue Fury devices and then calculated what I needed to set it to in CGMiner and did so accordingly; Likewise I accidentally forgot to turn off the MinePeon donations, so I saw my hash rate drop at the pool (ghash.io notifies me every 15min if there's a drop) and forgot that some pools dynamically set the difficulty, which BTCGuild does and it set it to 64, so I copied that setting.
Since doing all of the above, I haven't ran into a single LIBUSB error for over 3 days now...
(Yes, I did get them before doing that, about once every 7 hours on the rPi's and about every 15hrs on Windows7.)
Now, with that being said, stating that the product is faulty here, I'd declare that to be false; Maybe more so of people not knowing how to configure settings correctly, that may very well, be true. I'll let you all be the judge of that, but I've been mining on my rPi solid now. I think it's wise to factor in all the variables before making out right accusations that a product is faulty, especially when in mining, there are *a lot* of settings and configurations that most people aren't even aware of how to use. Are there improvements that can be made in the firmware to help this issue (maybe avoid it completely)? Probably, yes, of which Angus and the Drillbit team have stated they are looking into.
So, yea...