Here is what I have found works for me.
I use an up-spiral 2 flute .250 dia carbide bit. Straight flute bits tend to chatter with these hard materials. (and don't clear chips well either)
Depth of cut per pass is .02 - .03 inches ( yes, thats a LOT of passes) travel speed is 20 ipm and plunge speed is 20 ipm.
rpm is 8000-12,000. definitely not full router speed. The cut is done dry with no lubricant. all of the cutting is done at the tip of the bit, but tool life is still acceptable but not optimal.
YOU MUST RAMP THE CUTS.
Mostly the same here. 3/16 or 1/4" 2 flute. I never use a cutter in different materials, even if it is the same model tool - aluminum bits are dedicated to alum, wood for wood, plastic for plastic. The Atrax bits from Enco work well and are inexpensive, especially when they are having a 20% off sale on top of a bit sale.
I am uncertain whether a spritz of lube helps between passes. Whether in improved tool life, or surface finish. I tend to believe it improves the finish. An air jet may also help to cool the tool and prevent chip re-machining.
Regarding ramps, it depends on the type of cut. My most recent project required facing some mounting plates. In that case I chose to use a raster toolpath that resulted in all Z moves happening while outside the piece. So ramps were not necessary. I felt that approach was better than ramping on the piece. In situations where that is not possible, I agree that ramps are essential.
As you stated, with ramps your plunge rate must match the feed rate. If they are different, you risk abrupt changes in feed rate that can shake the machine.
The shark is very wobbly along the Y axis. When you move the Y, you can see it jiggle. You can feel it with your hand, or you can measure it with a dial indicator. So I try and avoid cuts that move the Y and instead leverage the greater stability of the X. So this often dictates raster toolpaths with the primary motion along the X. Also, in this project I extended the tool path far enough outside the workpiece that the Y gantry shake from step-over had stabilized before the material was engaged on subsequent passes.
The results in this particular project were excellent. This was just facing. The holes were made on a drill press. Covering the table with rags makes cleanup easy. Though I prefer the larger chips of aluminum over the fine dust of wood, but wood does smell better.