I’m extremely excited to finally see my latest paper with Davide Pedrotti, Luis Escamilla, Valerio Marra, and Leandros Perivolaropoulos, finally out on the arXiv - kudos to Davide who did all the heavy-lifting in what I do not hesitate to place among the three most important papers I ever wrote! This paper is about what we could call the “Hubble tension no-go theorem”, which forbids purely post-recombination solutions due to the constraints BAO impose on the product of the sound horizon and the Hubble constant. However, the pipeline from which BAO measurements are obtained requires assuming a fiducial cosmological model, with the choice falling on ΛCDM: many have therefore wondered whether we can trust these measurements when testing late-time modifications to ΛCDM, and this is the most frequently invoked loophole to the no-go theorem. In this work we have played devil’s advocate and showed that, even if BAO measurements were to be grossly (and unrealistically) biased by fiducial cosmology assumptions, this would still not be sufficient to rescue post-recombination solutions to the Hubble tension, primarily because of the extremely tight constraints on the (unnormalized) shape of the expansion history from unanchored SNeIa, whose role in the Hubble tension so far has been underappreciated. You can read our results in the preprint we just posted on arXiv: 2510.01974.
