At the very beginning layers should be eq'd separately like you said and you should adjust the level of the volume faders to get the desired balance. Experiment with the different levels of snappiness you can get by moving the clap before the snare. Once you get things to sound the way you want start off by merging the layers down to one audio file. What's the point of having 3+ snare/clap tracks if they're being treated as one sound? Condense them down to one and it will make mixing them 100x easier.

There is no need to use a drum bus. The key to getting a layered clap/snare to sound snappy isn't a drum bus. The key is experimenting with the amount of delay between the clap and the snare combined with the transient, decay, and EQing of the sample. You can adjust the shape of the sound by using either logics EX24 or compression.

Also just so you know even if you were to send your layers to a drum bus you shouldn't do it in the way you described. You wouldn't use a "send" because then you just have two layered clap/snare signals playing at the same time. You would assign them to a bus but not via sends. On the actual output of the track instead of using "stereo output" you would change it to the desired bus. THEN you would process all of them together as one sound via adding plugins on the bus.