Wilson's solution of using composites to make individual small enclosures is a good one. MDF is just as good you just have to increase wall thickness and keep the enclosures small with rounded edges and flush mount drivers.
Right. The problem with metals is they may look and seem strong to us while on a microscopic level they are actually vibrating and ringing like crazy. So they work great on one level because they are strong on that level. Being a lot harder than MDF they dampen macro-dynamics a lot less than MDF. Yet when it gets down to the microscopic level, well that is where the real action is because that is where you either get or do not get that spooky real degree of refined inner detail.
This is why the future of speaker design is like everything else composites. With composites the vibration control happens at the microscopic level. No longer a continuous medium like all the alloys composites are a, uh, composite of different materials. As such they get both their strength and ability to transfer mechanical energy into heat at the molecular level. Combined with multiple small cabinets custom shaped to optimize each driver, its hard to beat.
Except that like jon_5912 pointed out, it has to justify the cost. Designing with composites is extremely expensive, especially in the beginning when almost everything goes in the dumpster because you don't really know what you're doing. The first Porsche Carrera GT was supposed to be impossible to make profitably from carbon fiber. Then it was supposed to be possible but only in a $750k car. Then it came out for less than a third of that. Its still not common construction but the technology is filtering down, and the same is bound to happen in audio.