[Analytics] Fix bulk writer consistency level validations for blocked instances

The analytics library bulk writer spark job performs consistency level validations around write operations.

Currently, these validations incorrectly double count the number of blocked instances for these validations resulting in valid scenarios failing CL checks.