We do not have any undocumented SPI command which will cause the zeroing of Device ID and serial number. The Device ID bytes are Read only Bytes.
Do you see the same behavior with the new device also. On how many devices did you observe this issue.
Sadly, my sample size was 2. :) I haven't done much with the second sample chip so far, but I will continue to revisit it in the coming weeks.
I still can't rule out the possibility that the first sample chip was stressed inadvertently through ESD or incorrect power application, so I'm not ready to indict the chip design based on one failure.
The corruption may happen due to power ramp specifications violation of the device provided in the datasheet. This might be configuration register corruption which happens during the power up which is causing this failure. Please re-validate your power ramp transitions to avoid same failure on the other chip.