Several things in this topic...
I used "Quality" as the setting because that's what the reference encoder calls it. It's really an overall compression level setting. The higher the number the more compression -- but all of them are still lossless. I am with m1abrams in that I prefer to use 5 or 6 because it (subjectively) hits the sweet spot between compression and time. PS there is no way to tell afterwards what compression level the file is, so if you care, you should write it to a tag.
The "Quality" of the rip is based on the number of reads it takes to get good data. If you are getting logs with "2 reads required to get good data" then you still got good data, but the quality will not be 100%. If you are getting very many "unreliable data" entries, then you will probably be able to hear it. In a 16 bit file, your error may only be one inconsequentially inaudible bit. But you have to listen. And the drive you use makes a huge difference. There are many posts on the forums about the secure-ness and quality of MC's ripping engine. Using it in the secure mode makes it equal to EAC. Secure is definitely the way to go.
The use of "encode on the fly" or "rip and encode" is only whether a temporary WAV file is created. Depending on the compression level, the format being converted, and the speed of your system, one may be faster than the other. They also use different functions in the plugin, one which uses memory blocks, and the other which works on complete files. Functionally, they're equivalent, so use whichever is faster/easier. There is no effect on the "losslessness" of the file.