This is a difficult error to troubleshoot. I have a few suggestions.
Replace the data cables for each disk
Check to see if updated firmware is available for the drives
http://kb.sandisk.com/app/answers/detail/a_id/15108/If the same physical drive keeps falling off, I'd perform a Secure Erase on it before re-adding as a RAID member
Stripe size on RAID 0. Smaller vs large. I think you can find an equal number of posts for and against smaller vs. larger stripe sizes for RAID 0. I am a fan of a larger stripe 64k or 128K. This was more of an issue back in the day when we all used spinners. Now that SSD's are pretty much the norm, using a larger stripe provides best overall read/write performance in the majority of cases. Boot drive vs data drive. I suspect the majority of the data you are accessing is much > than 4k. Interleaving smaller data chunks between 2 or more disks adds greater overhead on the controller. It can handle it, but really doesn't need to work that hard. With the high level of performance you get from SSD there is not need to attempt manual optimization as the controller and SSD's will do this for you.
FYI
You cannot randomly change stripe size on the fly and test read/write performance. You would have to restore the entire array at each iteration. Also, ***please back up your data regularly. RAID 10 does not mean you are 100% protected.