The paper reviews recent SOLPS modelling of long-legged divertors carried out by UKAEA and the University of York. The required upstream density required for detachment is predicted to scale inversely with the total flux expansion of the divertor, but experimentally this was not found to be the case on TCV. Interpretative SOLPS-ITER modelling suggests this to be a result of the confounding variable of neutral trapping; when neutral trapping is artificially equalised in the simulations, the effect of total flux expansion is as expected. Analysis of predictive MAST-U Super-X simulations suggests that the detachment front location in parallel space becomes less sensitive to control parameters (i.e. more easily controlled) in regions of high magnetic field strength gradient. The same simulations also suggest recombination sinks to be present only in deeply-detached simulations with relatively high upstream density and low impurity fraction (driven by deuterium fuelling). In equally-detached simulations with lower upstream density and higher impurity fraction (driven by impurity seeding), recombination was negligible. Deuterium ion-molecule elastic collisions are necessary for this qualitative difference. Finally, the relevance of long-legged divertors is investigated on DEMO. The engineeringly-feasible increase in outer divertor connection length is found to significantly increase the operational margin compared to the standard ITER-like DEMO divertor, approximately in line with simple Lengyel Model estimates.