The phenomenon of self-absorption is by far the largest influential factor in the efficiency of luminescent solar concentrators (LSCs), but also the most challenging one to capture computationally. In this work we present a model using a multiple-generation light transport (MGLT) approach to quantify light transport through single-layer luminescent solar concentrators of arbitrary shape and size. We demonstrate that MGLT offers a significant speed increase over Monte Carlo (raytracing) when optimizing the luminophore concentration in large LSCs and more insight into light transport processes. Our results show that optimizing luminophore concentration in a lab-scale device does not yield an optimal optical efficiency after scaling up to realistically sized windows. Each differently sized LSC therefore has to be optimized individually to obtain maximal efficiency. We show that, for strongly self-absorbing LSCs with a high quantum yield, parasitic self-absorption can turn into a positive effect at very high absorption coefficients. This is due to a combination of increased light trapping and stronger absorption of the incoming sunlight. We conclude that, except for scattering losses, MGLT can compute all aspects in light transport through an LSC accurately and can be used as a design tool for building-integrated photovoltaic elements. This design tool is therefore used to calculate many building-integrated LSC power conversion efficiencies.