My emerging theory over the past decade or so is to avoid anything that is rigid to the point it can ring in any way. This goes for many things, e.g. bolting an arm to an armboard, a driver to a baffle, pretty much anything. I no longer have any low-mass speaker stands to play about with, but my feeling with high-mass ones is not to over-tighten the bolts, do them up to the point they neither rattle nor ring. Basically always avoid turning your hi-fi into percussion instruments. With my speaker stands for the LA3/5As or JR149s the spikes just push through the carpet under their own weight (heavy stands) and are then levelled, I don’t use a spanner, just finger-tight is fine. The stands themselves are tightened by ear, i.e. no rattles or resonances.
The idea of trying to transmit as much energy from the speakers into the floor (typical ‘80s attitude) is hugely flawed IMO as a floor itself is a resonant structure, and an unpredictable one at that. Far better to try and lose energy wherever you can do so without generating resonance. A 5kg speaker on a 20kg stand is not going to be moving about much anyway at audio frequencies, it is just too high a mass for a little driver moving at say 50Hz to 20kHz to set in motion, so I’m pretty convinced what you want to do is to stop it either having an audio frequency resonance of its own or moving a hollow and potentially resonant floor. If I was in a non-carpeted room (a disaster sonically IMHO, so I never would be) I’d certainly try experimenting with compliant interfaces. No way in hell would I pay Townsend or equivalent prices, but I suspect there is something to that logic that is worthy of exploring. There will certainly be industrial equivalents.