gary, here is a quick breakdown of the s3d support generation.
the support system in s3d is based on a grid. it looks like a piece of graph paper. the size of the squares are set by the "support pillar resolution". 2mm is a good size for an average print. what happens is when a support grid square becomes covered a certain percentage then that square is activated and a pillar is produced. if you have a fine detail part with an overhang that sticks out but doesnt cover enough of the grid to activate that square then that is when you have an issue with unsupported overhangs. you can correct this by lowering the resolution of the grid and turn on a setting called "extra inflation distance". 1mm seems to be a good middle of the road setting with this. inflation distance takes the generated support and expands it outward basically covering those areas that overhang but arent large enough to activate the pillar. hope this makes it understandable.
support infill % is the same thing as your regular internal infill % except it applies to the support. the higher it is the more dense the support structures will be. the grid pillar resolution has no effect on support density. that is only to configure the location and area the supports will be printed in.
for question 1.....i dont think there is a setting for that.
as for dense support layers, s3d will now allow you to print a low infill support say 15% then you can choose the amount of layers at the top of the support (we will say 3) that you can print at a higher infill %...say 60%. now you have a quick printing, low plastic using support while still getting a 60% infill at the top 3 layers. this makes for nicer bottom layers on overhangs.