Unfortunately the root cause (the FATAL error because of writeable check failure) cannot be solved with the code, it need to be addressed in your setup.
No, we do not have such plans.
if you are using one node per disk, you can improve the I/O throughput a bit:
perform a defragmentation and enable the automatic one (it’s enabled by default);
disable 8dot3:
disable atime:
The next step would be to enable a write cache on the drive, but it requires UPS to be safe.