Support parallel simulation of multiple sources

Use the power of numpy advanced indexing and broadcasting to allow an arbitrary array of input sources to be simulated in parallel, with minimal changes to the code. This is presumably useful for simulating the observation of an entire field of targets, and eliminates the overhead of looping over each target. The existing mode of simulating a single 1D input SED should still work without any change to user code.

Start by locating the existing single 1D input SED simulation entry point and its tests. Check how NumPy advanced indexing and broadcasting can accept arbitrary source arrays while preserving the current user-facing call, then verify both single-source and parallel-source behavior.

Written by the indexing model from the issue text.