Hi Monica (and Jonas),
Monte Carlo simulation is so far not possible for projects, because it is computationally somewhat heavy. For a comparison of e.g. two products, create a product system that has as central process (that is providing the quantitative reference) product (e.g) conventional rice as input and product organic rice as -input (so, +1kg conv rice, - 1kg organic rice, output and quantitative reference if the process is "something" - doesn't matter, e.g. rice, 1kg)
With this, create a product system and perform a Monte Carlo simulation. If results are mostly negative, conventional rice is better (organic rice has higher impacts), and vice versa.