Does the dish on a rear wheel really have that much affect on the strength of the wheel? Or is it really that much of an issue?
The old 5 cog freewheels had their own problems. But, I guess you could get a more neutrally dished wheel by squashing 5 modern thin cogs onto a short freehub body where the hub has more distance between the flanges. The wheel design book I have (schraner) makes no mention of dish having any affect on wheel strength. What has more affect is flange diameter, rim depth, spoke length, spoke count and cross. And you can do (and companies have) done funky things to mess with the first two. Deep section rims, asymmetric section rims and asymmetrically flanged hubs being just a few examples.
From what I've read the going theory is that spokes are the weak link in the system. You reduce that weakness by making the weak part shorter, and by causing there to be more weak bits to take up the load (more spokes is generally recognized as being stronger than fewer spokes). What exotic wheel companies have done, more often than not, is produce wheels with fewer spokes by making the rim stronger and heavier.
On the flip side, the average rider only ever uses 5 or 6 of the cogs in a 7 or 8 (or even 9) cog stack. Which IS a waste. But, the industry can't sell hybrids that don't have granny gears even if that granny gear never gets used. However, the argument for having tiny steps across a short range of gears holds on the fact that the work you put in a bicycle system is more efficient if you allow your engine (ie legs) to run at a more constant cadence while changing gears. For a demonstration of this, hop on a bike with the Shimano automatic shifting system (Trek Lime, Raleigh Coasting etc) and go for a pedal. It is hella jarring when that system shifts into a higher gear.