As has been said a dozen or so times - no it won't make a difference in terms of audibility; we're talking about a very small jump. What I find somewhat comical is that the debate over "well halve the power - thats reality" goes on after it's been stated that even double power doesn't make a difference (3 dB and all).
I've seen a number of reviews that substantiate receivers in the $1000+ range delivering their rated output (or close enough as to not be worth arguing over), and I've seen similar reviews that show $300-$400 products falling off a cliff.
Some examples:
http://www.hometheater.com/content/s...-labs-measures (what we absolutely don't want to see - there's so little power available that headroom does actually go out the window in this situation; 1-3W input will probably get you the 80-90 dB output, but what about the 10 or 100 fold power increase for dynamic peaks?)
http://www.hometheater.com/content/s...-labs-measures (what we absolutely do want to see)
http://www.hometheater.com/content/i...-labs-measures (what we absolutely do want to see)
http://www.hometheater.com/content/a...-labs-measures (for the price, what we absolutely do not want to see)
http://www.hometheater.com/content/p...-labs-measures ("average" - this is roughly what people are getting at in terms of midrange equipment)
http://www.hometheater.com/content/s...-labs-measures (actually outperforms more expensive equipment in terms of ACD - cost/brand name can't always be used as an indicator of performance)
http://www.hometheater.com/content/y...-labs-measures (one of the actual units in question - hitting fairly close to its targets)
Now, Yamaha specs that thing at 100wpc, and we're talking about 30W of difference, so what's "lost" between the 100wpc claim and the 70wpc output? About 1 dB.
As to why this doesn't matter, at least not as much as we'd think it does, let's continue. First off - in order to perceive a "doubling" or "halving" of loudness, the signal's intensity has to increase or decrease by 10 dB - not 1 or 2 or 3 dB. Secondly, all-channels-driven is a maximum output scenario, music and movies won't ever re-create this (if they did, there'd be none of this headroom that's being argued over) - so if we read a bit further into that Yamaha's numbers, it blows right past 100W into a single channel (170.5W). If we figure that a single channel, or perhaps a pair of channels, will actually be demanding peak outputs at any given point - well, I can't say this absolutely, but I'm guessing you'll get that full 100W (assuming the signal is "hot" enough to demand that).
Not all receivers lie about their numbers, and even if the device doesn't reach its all-channels spec, it will probably deliver that power into a single channel (even the godawful Sherwood gets pretty close at around 75W, close enough that I doubt you could pick it apart from most of the other units listed up there unless you were doing 7-channels ACD against one of the Sony units, for example). This is probably "good enough" for most situations. That said, I'd probably take something like the DN1000 over the VSX1120, or the Integra over the Anthem.
Finally, using the nameplate rating on the receiver is potentially an inaccurate way to measure its power supply capabilities - often this is an average consumption value, or some invented number. Remembering that the average user will probably never need more than a few watts per channel, the AVR itself is unlikely to consume much more than what is needed by the internal DSP devices. This means that even the fanciest, highest end units can get away with 300-400W consumption values, even if absolute maximum input approaches something like 1000W. If you're curious about how much power is going into your receiver, get a wattmeter and test it out - I'm sure at normal volume levels it probably doesn't budge much over what the computer parts need to run, but if you really crank it will probably draw quite a bit of power from the wall (just like any other electronic device - it only consumes what it needs to operate plus loss).
Let's use the Integra from the above links as an example - on the back of the unit it states 7.5A (here's a nice huge picture:
http://www.gspr.com/integra/images/dtr_502_rear_300.jpg ), which is 900W from 120VAC. In seven-channel output mode it posts 127.7W at 1% THD, which is 893.9W. If that nameplate rating were true, the Integra would be 99.3% efficient; not even the fanciest Class D amplifiers in the world on their absolute best day could approach that. Reality is, that unit is probably drawing something in the realm of 1500-1600W or better to accomplish that output. We can look at another unit that performed fairly well, the Sony DN1000, which has a nameplate rating of 250W. That unit provided 64.9W at 1% THD, which is 454.3 watts - if the nameplate rating were true the Sony would actually be generating free energy, and have an efficiency of 181% (wouldn't that be nice!). This also assumes the DSPs and other components in these devices draw absolutely no power. However, remember that neither of these units will ever see this scenario aside from a testing situation, they're likely to see 1-5W per channel continuous average, and peaks of 50-100W on a single channel periodically (if ever, we are assuming fairly high outputs are desired) - add that consumption plus loss due to inefficiency to the power drawn by the DSP components, and you can easily fit into the nameplate ratings.
Some further reading about the perception of loudness; amplifier output, sensitivity, and power; and SPLs (and, much to Michael's dismay, because it was hinted at, damping factor):
http://trace.wisc.edu/docs/2004-About-dB/
http://audiokarma.org/forums/showthread.php?t=58829
http://myhometheater.homestead.com/splcalculator.html (you can figure out theoretical output based on speaker efficiency and amplifier power with this toy)
http://www.physics.mcgill.ca/~guymoo.../lecture11.pdf
http://www.butleraudio.com/damping1.php (this is essentially identical to the Roger Rusell website, but presented as text instead of a pdf)
As an absolute addendum (and I think the nails are pretty well in at this point) - the "power freed up by the subwoofer" argument is another curiosity. If we remember that the equal power frequency sits at something between 250 and 350hz (in other words, half of the amplifier's power is needed above and half below this frequency), that would mean if a subwoofer's crossover point were somewhere in there (let's say 300hz because it's comically high and right in the middle), it would remove "half" of the power demand from the other speaker channels/amplifier (this assumes an absolute brick-wall of a crossover, iirc THX says 12 dB/Octave is standard - that isn't exactly a brick wall, but it isn't exactly a rolling hill - you can do the maths and figure out the numbers more exactly though). That "half" would represent 3 dB of additional headroom (so if you went from a hypothetical 30W/ch to 60W/ch), which takes us back to "what's audible?" (10 dB) - see the problem? If we go further into the power distribution across frequencies "thing", the point where you'd see 10-fold reductions in amplifier demands is somewhere around 5khz - an entirely reasonable crossover point for...a midrange driver. Now, we are talking about averages and generalizations here - if we take other factors (like acoustics) into account, low frequencies should be more intense in order to be as loud (and low frequency peaks can demand large amounts of power as a result of that and other factors (driver efficiency being a factor too - your average subwoofer driver isn't going to be as efficient as your fancy horn tweeter)). Also remember that content isn't a test wave (everything is lower than full-scale). Ultimately though, even if the subwoofer could somehow remove 50% of the power demand from the amplifier (which it generally will not/can not), we're still back to 3 dB of additional power ("60W from 30W") - not 10 (300W) or 20 (3000W) dB. I'm not arguing against subwoofers, but am arguing that adding a subwoofer to something like the godawful Sherwood will not help it out in terms of somehow making it able to deliver enough power to bring it to the level of the Integra, Yamaha, or Sony ES boxes.
See here for more:
http://sound.westhost.com/bi-amp.htm#power_dist