The concave shaped open/close ramps of the OEM cam lobes are said to cause high acceleration forces to the valve train. This topic has been thoroughly discussed here at UHE. It is likely that these forces are shortening the life length of the cam roller bearings. Dale Lineaweaver is providing a remedy to this with his LX1 and LX2 cams.
I need larger duration, higher lift, lobe centers further apart and less acceleration forces, which has led me in the direction of trying to produce a cam of my own. I will run the profiles #62, #53 and #08 in a "Cam Doctor" rig, to understand what is going on.
In my opinion, the conical springs' only benefits are like you say, less mass and less space required compared to a cylindrical dual spring. The conical spring gives resonance over a wider spectrum and offers no redundancy when a coil breaks.
To be able to say anything about the risk of using dual cylindrical coils together with steel valves, you need to test the springs at installed length and at max lift with a spring tester. Most well sorted tuning workshops have one.
I have not gained enough experience of my own yet, so I'd better keep my mouth shut for now.
:-#