Okay. I had some time this morning so I gathered every USB drive I had that had 1 GB free on it. Ended up being 11 different ones - christ I didn't realize how many USB sticks I had. I then benchmarked how long each one took to transfer 735mb of data to and from. Pretty much each drive was different. Everything from Memorex to Sony to Sandisk to Lexar to random unnamed ones (most of them are ones I found discarded while I was working at a university, some I've bought from various places and for various reasons). And I decided to try and let science and observation settle this. Though in saying this I used my macbook only so perhaps that slants the findings as I know nothing of how the architecture of the xbox works with USB drives. But in transferring a file I could not find a single difference between any of the drives. They all transferred a 735mb AVI file to my macbook in and around 44 seconds and it took quite a bit longer to upload it to the sticks but each one took around 2 minutes and 20 seconds, though perhaps in retrospect that might have also been due to the fact that I was making it overwrite an existing file. Maybe I should have done a clean upload but still. There are my results. I stand by my original statement and feel far less ignorant in offering this advice now. According to everything I can see there seems to be absolutely zero difference between any of these drives in terms of speed in terms of file transfers. Perhaps there are drives out there that are better or worse than others but it seems typically that this is not the case, that they seemingly are built in similar ways and provide similar results. Though as I said I don't know how the Xbox architecture works and I didn't run any programs off the sticks to test speeds so perhaps my test is moot. I don't know. But my feelings is that they are all pretty much the same.
No. There is a lot lot lot more to just timing the transfer speed of a single large file. It's okay if someone doesn't know how it all works, but some small sampling of anecdotal evidence is not scientific. Reading large sequential files are the fastest. It's a bunch of small random writes that are slow. If you check prices, you will see USB 2.0 flash drives that offer really truly fast read/write speeds, but especially write speeds, cost a lot more than these el cheapo drives everybody has in their collection. Testing a bunch of low cost flash drives tells you that they are all similar, well of course.
In the long run this stuff only matters if the USB drive gets heavy use everyday, it's unbearably slow, and you are willing to pay more for better performance.
To get an idea of how it works, try the program 'CrystalDiskMark 3'. You can use it on any kind of drive. You will have to do some googling and wikipediaing to understand what all the different numbers mean. In essence, if the USB 2.0 flash drive has the typical 4MB/s sequential write speed, it's a slow one. It should be 10~20MB/s write speed for good performance.