This paper presents our recent developments that aim to make the instrument definition process of our visual synthesis environment more accessible to a broader audience. There are several novel aspects that aim to overcome some of the classical limitations found in sound synthesis systems. After an introductory section we discuss two advanced examples where scores containing several model-based instrument parts can be realized without having to edit the original instrument definitions. In the first one, we can duplicate instrumental parts of a single instrument definition. In the latter example, we can mix in a single score several instrument models with the help of a mixer patch that is automatically created by system.