Disk Utility's secure erase options do not read back the zeros (or other data patterns used in the multi-pass options), as the user tip incorrectly claims it does. (The same is true for the diskutil erase options.) Thus, this step doesn't map out bad sectors, which can only be detected during reads by the drive itself.
The only function of a secure erase is exactly what the name implies: to securely erase any existing data on the targeted part of the drive, which is done by by writing over it with one or more meaningless data patterns. It is not necessary to read anything from the overwritten sectors to do this; in fact, since reading a weak sector during the process might cause the drive to map it out while it still contained recoverable data fragments, it is undesirable from a security standpoint to do this.
Plus, from a practical standpoint, reading back the sectors would more than double the time the secure erase would take & cause pointless work for the drive. Writes take less time on the average than reads because the drive automatically performs integrity checks on reads, which sometimes requires more than one read attempt even from good sectors.
If you doubt that a secure erase is a write only process, you can check this for yourself by getting the drive's specifications & comparing its read & write performance with how long it takes to do the secure erase. If it takes less time than it would to both write & then read the sectors then you know it can't possibly be reading back anything.
