One basic reality is that many relevant specifications are not published for many amps, and if figures are given, they're often not specified in ways that can be compared directly. A lot of shady speaker amps will fudge the output power numbers, for example. A lot of metrics will change significantly depending on the impedance of the load (headphones) used and the output level, not to mention the frequency of the signal sent, so you need equivalent testing conditions to do fair comparisons between different amps.
Some interesting articles are archived here about audio:
http://gilmore2.chem.northwestern.edu/tech/index.htm
Basic information about different audio performance metrics that may be used for specifying amplifiers:
http://en.wikipedia.org/wiki/Audio_system_measurements
InnerFidelity posted an article about headphone amp specifications and guidelines about what is good, but there are differences of opinion regarding what is "enough" and so on, at least for some of the parameters. And these are just guidelines to what is more accurate (higher fidelity). Even though lower output impedance is technically "better" that does not necessarily mean that you will like the sound better with lower output impedance, with some or many headphones. The same goes for some other parameters. See here:
http://www.innerfidelity.com/content/nwavguys-heaphone-amp-measurement-recommendations
