Jeff,
Definitely wasn't me. While I find that spectrum width has little added value in normal supercells given the storm scale variability/turbulence in velocities, there is potentially significant added value using spectrum width in a hurricane where you're dealing with a more "smooth/laminar" flow so to speak, and interruptions like a tornado could show up fairly easily depending on the supercell characterstics and proximity to the central core. Most hurricane spawned tornadoes form in mini-supercells in the outer rainbands in the N and E quardrant. Tornadoes or meso-vorticies also form in the inner eyewall of intense hurricanes and can lead to significantly enhanced damage in small areas (this may have been the case yesterday).
If you remove the forward speed of the hurricane by using SRM instead of base velocity, I have found that "tornadic signatures" are much more easily identified. We identified some pretty heafty couplets ahead of both Bonnie and Charley, many of which produced tornadoes, and I was impressed by the lead times offered by the Florida NWS offices given how fast these vorticies move (often forward speeds of 50 to 70 mph).
More often than not, your ingredients for tornadogenesis in these rainbands are already present in the NE quadrant...enhanced low-level storm relative helicity and a warm/moist RFD. Therefore with sufficient instability present to support a rotating updraft (i.e. clearing between the rainbands), I've found that any storm usually able to reach 55-60dBZ, last for 30 minutes and reach a modest depth has a very good chance at producing a brief tornado.
Evan