Thank you very much everyone, a lot of helpful and informative comments!
Typically our samples do not have low mass absorption so this advantage of parallel beam is probably not so important for us. We do aim for high resolution, even with the time cost, and new solid state detectors will help. Still, I guess intensity loss from a mirror & 2-bounce monochromator will be more than from Johansson monochromator. As suggested, comparisons of data and modelling against it from both geometries is really needed. Thanks again! Markus