In order to infer the BAO along the history of the Universe, it is necessary to pinpoint the baryonic large scale distribution, and this has been achieved mainly via the stellar content of the galaxies, as it is the case of BOSS project, or the pioneering work with DR3-SDSS.
In a more general approach, the most direct way to detect matter should be through the most common element (within 5% of “normal” matter) in the Universe … which is atomic hydrogen.
HI is all over the Universe. The problem lies in the fact that atomic hydrogen is extremely difficult to detect because it radiates via an extremely improbable phenomenon: a photon is emitted when the electron and proton spin changes from a parallel configuration to an antiparallel configuration, occuring a spontaneous transition each 10 000 000 years ! The energy of the radiation corresponds to λ=21 cm or ν=1.421 GHz in rest frame. Because it is so improbable, to be detected either the emitting region is very extent or the collecting area is large — as it will be Square Kilometer Array (SKA) interferometer .
BINGO is about a third and somehow intermediate possibility: to measure HI on a large scale with a modest telescope in terms of collecting area, using a new technology called intensity mapping. When mapping intensity, all the photons, (in this case hydrogen) count, the S / N not being relevant (can go as low as 1).
Here, the main motivation is to identify the BAOs, and for this it is only necessary to identify the large scale fluctuations in the distribution of HI without the need of differentiating the individual sources, that is, the galaxies.