Distributed lighting builds will most likely be limited to Enlighten in the first iteration. In Enlighten the jobs are a somewhat heavy and all the inputs are normally read from disk. This works really well with the initial approach to distribution that we took.
The progressive lightmapper is more lightweight, each job needs to finish quickly, but at the same time build upon the results of the previous one. The approach to distributing it will be different, most likely involving a bunch of live contexts solving a (part of a) lightmap each. It’s just a different… beast.