DPMS and HD power saving are nowhere near as problematic as sleep and hibernate.
Although, in all honesty- my new laptop has had zero issues with any power management... well, except when I ran XP on it, since it woke up from sleep and the processor was still clocked down from 2Ghz to around 16mhz... but in Vista and 7 it's been flawless. (installing 7 apparently reset my power options, so I went to use it and instead of simply being at the log-on prompt It had hibernated, and it resumed just fine).
Basically, DPMS is pretty simple. It doesn't actually "power off" the display completely, but it stops sending a signal to the monitor; most monitors use far less power this way, and more often then not go into a "sleep" mode themselves until a signal is restored.
Hard drive power options are just as safe, as long as the drive is in good working order. After a period of inactivity- the drive motors power off. drive motors are one of the biggest consumers of power in a PC, being the main draw on the 12V from the power supply. The drawback is, of course, that before reading or writing from the drive again the OS has to wait for the drive to power up. (well, actually, the OS doesn't wait at all, the HD controller circuitry ON the drive simply refuses to return data until it can actually read if) so, if the drive takes WAAAAY to long to spin up, then you might get a "delayed write failure". However, if the drive takes that long to spin up (usually in the order of >10 seconds) then the drive has deeper issues that should be addressed, and is likely facing an impending failure.
my Thinkpad 755CDV and Toshiba 440CDX both power down the hard drive after a period of inactivity; and it seems that while I can influence this through the provided power options I cannot shut it off altogether without visiting the BIOS. This is due to the architectural design of the original power management specifications- APM- APM defines BIOS extensions, but makes no allowances for software control over those BIOS functions; basically, APM defines that the BIOS will alert the software when the PC is going to sleep but will completely ignore whatever the Program does; basically, it says, "OK, I'lm going to sleep now"... "but... but... I'm saving a file!!" " Too bad. sleeping".
This architecture often resulted in data loss. This was the main reason for the creation of ACPI, which is essentially a redesign of the original APM specification to more closely integrate the software and operating system into the decision making processes that were originally left squarely up to the BIOS.
