Wow, zombie thread.
To those who think this is a dumb idea, consider this -
When we hear the low E of a standard bass guitar, most of what we hear is not the fundamental (41Hz). Sure, there's some of that there, but most of what we hear are higher overtones - primarly 82Hz. But we still percieve this as a "low E".
Note that this is also why the iconic Ampeg 8x10 can have a slammin' loud low E, while at the same time this cabinet been measured over and over again to have a frequency curve that rolls off rapidly around 60Hz - well above the fundamental of a low E.
So, assuming the same performance from the instrument as we go lower, I would also assume that playing a string tuned an octave lower would produce mostly an overtone around 41Hz, with a bit of fundamental in there and some higher overtones.
41Hz is well within the measured performance of many "hi-fi" bass cabs and most decent PA systems. So is this really such a dumb idea?
Since, in this example, you're still talking about sending some level of a 20.5Hz signal, I would say it'd probably a good idea to be careful about hitting even the best systems with frequencies that are below the tuning frequency of the bass cabinet. This is where a speaker becomes "unloaded" - i.e. starts flapping around like a jackhammer and blows pretty quickly. So keep an eye on your speaker cones and figure out the limits of the system you are playing through. You'll want to do this kind of experimenation BEFORE a gig, before rehearsal, etc.
Even with proper low frequency protection, you can still get a lot of perceived low end, so it's worth trying.
Is it useable? That's up to the individual player to decide, but if anything is "dumb" here it's declaring this kind of experimentation to be useless. If you're going to make judgements based on theory rather than actual experience, you'd do well to actually think through your theory first, rather than post up a knee-jerk reaction.