question to subprocess grouping

Hi,

I'd be curious if the subprocess grouping is applied in the following case:
generate p p > e+ e- e+ ve @0
add process p p > mu+ mu- mu+ vm @0
add process p p > ta+ ta- ta+ vt @0
I have set all leptons to be massless before, so the grouping should be in principle possible if I'm not mistaken.

Here are some background information:
I'm interested in producing WZ events in the fully-leptonic final state, i.e.: p p > l+ l- l+ vl with 0,1,2,3j @LO using CKKW-L kt-merging. I need to apply a mmll cut, to cut the photon divergency, which isn't possible when I just use "generate p p > l+ l- l+ vl". My plan was to split the sample and calculate the same flavour and the opposite flavours for each W-boson charge independently from each other (e.g. e+e-e+ vs e+e-mu+ vs e+e-e- vs e+e-mu-). So ideally, I could use the process definition as mentioned in the beginning to allow for the mmll cut while grouping as much subprocesses as possible. Would there be another way to separate the same flavour from the opposite flavour channels?

Thanks a lot,
Carsten