I recently read a posting from Privatex written last year making foam cones not out of the material from sanding blocks, but from high-density polyethylene or HPDE. This is the foam used for swimming aids or the mats for yoga etc.  So I replaced my snare and 1 tom with the new cones and what a difference. So replaced the lot.
They are much more sensitive carrying the vibration from the  skin to the piezo which greater clarity.  Far far easier to make and being more dense you can make the tip smaller alleviating hot spots, seven took me less than half an hour.  
It was suggested to glue a small tip of the old style foam however this seems not to be necessary. The foam came from a yoga block bought from a sports store, 4" x 3" x 8" long and costs £1.99, enough for 20 cones and being bright pink shows up under the white skins.
To make them cut a 1 1/2" cube cut roughly to a cone shape with a scalpel and then stuck to a cutting disk on a dremel with double sided tape.  Using reasonably rough paper and starting from the bottom work you way to the end leaving a 1/4" tip.  Support and balance the cone with your finger to counter the paper, however do not press hard with either let the paper do the work. Finish with a lighter paper although not necessary.  This type of foam also sticks better to the tape.
			
		

 By far the best result I've had ever. When I used harder foam with a small tip, the mesh tended to vibrate more, leading to more double triggering and forcing me to raise DynLevel and DynTime, and the softer foams tended to be a bit too insensitive. This has more surface contact with the mesh but there is barely a hotspot and the bigger contact area dampens the vibrations of the mesh, allowing me to lower Threshold, DynLevel, DynTime and Minscan. I tried with and without a disc the size of the ceramic part underneath and found it is really necessary to let the piezo flex. The only downside is that PS doesn't work as well, but to be honest I don't really miss it that much - I prefer to have more accurate roll detection and dynamics. And this one smiles at you
 By far the best result I've had ever. When I used harder foam with a small tip, the mesh tended to vibrate more, leading to more double triggering and forcing me to raise DynLevel and DynTime, and the softer foams tended to be a bit too insensitive. This has more surface contact with the mesh but there is barely a hotspot and the bigger contact area dampens the vibrations of the mesh, allowing me to lower Threshold, DynLevel, DynTime and Minscan. I tried with and without a disc the size of the ceramic part underneath and found it is really necessary to let the piezo flex. The only downside is that PS doesn't work as well, but to be honest I don't really miss it that much - I prefer to have more accurate roll detection and dynamics. And this one smiles at you 
