Crompton et al. (2023) developed a new algorithm for describing the performance of coarse particle flotation. They used the flotation rate constant, k, normalised by the maximum rate constant, kmax, for the pure mineral, as a proxy for the fractional surface liberation. The algorithm was used to produce the partition curve for a separation performed by a novel device, the CoarseAIR™. Part I of this new study re-visits the former work, particularly the batch mechanical flotation responses of the steady state samples from the CoarseAIR™. The flotation responses were deconvolved to the corresponding distributions of rate constants for the three streams, and in turn used to produce the partition curve for the coarse particle flotation. The algorithm used to produce the distribution of rate constants was driven towards a simple functional form by minimising its overall curvature. The steady state samples from any coarse particle flotation system can be assessed in this way. Part II of this study focuses on the reproducibility of the approach, and hence the uncertainty, using a batch mechanical cell to simulate the coarse particle flotation, and in turn the steady state feed, product and reject samples.