Pivot, read ANY review that includes true measurements and you'll almost always see something along the lines of: "In reality it was throwing out more like...."
The amp/brand then gets praised for rating their power specs conservatively, but it almost aways means that the whole doubling inbto 4 ohms goes out the window, as the 4 ohm figure is usally much closer to the manufacturer's specs.
I have NEVER seen measurements of an amplifier that actually completely doubles its output power into 4 ohm when measured by someone else than the manufacturer, in reality the 8 ohm figure is always higher than it's rated, sometimes just a little, but quite often by as much as 50%.
Example: Mark Levinson no.27.5 I owned was rated as 100-200-400 watt @ 8-4-2 ohm, in reality it measured more like: 150-250-400 watts. I've had amps from Krell to Classé do exactly the same on the repairman's test bench: The real life specs of an amp usually look much closer to the specs Japanes companies used to provide: Output power +50% into 4 ohm , something like 100watts into 8 ohms and 150 into 4.
Maybe Mcintosh is a brand apart, but I'm assuming those benchmark tests you had to perform were to ensure it MEETS its specs, not if it exceeds them.
Now assuming manufacturers do this to look good might seem cynical to some, but I can't think of any other reason they would knock as much as 25%-50% of their 8 ohm power specs. Particularly in an audiophile climate where the mantra seems to be: "Output power is not the most important spec: You can see if an amplifier is "stable" when it doubles it's output power into 4 ohms"
And why else would a manufacturer decide to "conservatively" rate his amp at 8 ohm, but go for an exact measurment with the 4 or 2 ohm power specs?
You dont have to take my word for it, stereophile has measurements in the end of their reviews, just look up your favourite power-house amp to see what I mean: 25-50% deviation from the manufacturers figure in the 8 ohm specs but only 5-10% in the 2ohm spec are the rule, not the exceptiom. Here's the Levinson 23.5 (they didn;t test the 27.5) http://www.stereophile.com/content/mark ... asurements
I am not saying it matters for the quality of the amplifier in the end: A Mark Levinson no. 27.5 (Or Krell, or Classé) is still a killer amplifier able to drive pretty much ANY speaker, even if it doesn't double into 4ohms like the manufacturer says. But "conservative" power output specs are often seen as a redeeming quality: A form of honesty and carfullness from the manufactured while they could just as well be a devious sales pitch to make the manufacturer look good. (In 2 ways as it turns out)