It's not overclocking the CD drive at all.
Many companies will use rebadged parts over a series of models and they'll put firmware on them to govern their speed. For example, when HP released their 8250 drives, the first batch of drives were basically Phillips drives with a different face plate and firmware. But these drives were prone to failure, which caused HP and Phillips to enter into a legal battle.
In the meantime, HP went ahead and continued releasing the 8250 drives but instead of using the Phillips internals, they went ahead and used the exact same internals from their 9100 models - which were nothing more than rebadged Sony CRX-140 E drives (8x4x32). To match the specs of the 8250, HP installed firmware that governed the speed down to 4x4x32.
In other words, anyone that bought an HP 8250 drive around March/April of 2000 were basically getting an HP 9100 with scaled down firmware.
But HP screwed up. Not only did this firmware act as a governor, but it also disabled the PnP ID chip on the drive, a chip that identified the drive as an "HP CDWriter 9100+" - the true nature of the beast!
That way, when the PnP BIOS detected devices, it had to poll the firmware of the 8250 instead, and that firmware would display "HP CD Writer___8250i". So, in essence, the 8250 was more than capable of 8x4x32, but the firmware HP installed governed it down to 4x4x32.
Some gutzy people on Usenet tried an experiment - they compared their respective HP 8250 drives using their original box and other credentials and found some consistancy in knowing that all their drives were exactly the same and were really close to the model numbers of the 9100. They went to HPs website and grabbed the latest firmware update for their 8250 drives and also grabbed the firmware for the 9100, as well.
While flashing the drive with the 8250's update, they'd power off their machine mid-way through, purposefully corrupting the firmware and rendering their drives useless. The typical green lights on the faceplate turned red, indicating a serious error - the firmware was toast...and the PnP ID chip was now enabled.
When their motherboard BIOS called for the PnP devices during POST, they saw the drive being identified as, "HP CD Writer 9100+" instead of an 8250. That paved the way for them to use the 9100 firmware update to get their 8250 drives back in working order...and with the bonus of an extra 4x for FREE, but at the cost of their warranty with HP, of course!
Usenet exploded. Tons of people looked at the credentials to see if their 8250s were internally identical with the 9100/Sony CRX-140-E drives and, if they determined that to be the case, they went on a flashing spree turning their 8250s into 9100 (keep in mind that, in March of 2000, the 8250 was selling for $149 while the 9100s were $299).
I was one of those people.
And my "hacked" HP 8250 is going strong today, although it is awfully long in the tooth compared to the 40x TDK VeloCD I saw at Wal-Mart for $109 a week ago!!
So, technically, it isn't overclocking - it's just firmware hacking.
Many CD-RW drives on the market today are just rebadged Ricohs, Sonys, and Plextors, etc., with governing firmware that lowers their speeds down from their REAL internal specs. HP, however, put a stop to the firmware hacking on the 8250 by using different internals by July of 2000, making it pretty difficult to find the March/April models. I think these drives were labeled as HP 8290s.
Edit: Most LG drives are actually re-badged Ricoh drives without some features and with a "governor" on the firmware.