Cosmography and flat $\Lambda$CDM tensions at high redshift

Risaliti, Lusso \& collaborators have constructed a high-redshift Hubble diagram of supernovae (SNe); quasars (QSO) and gamma-ray bursts (GRB) that shows a "$\sim 4 \, \sigma$ tension with the $\Lambda$CDM model" based on a log polynomial cosmographic expansion (Risaliti et.al 2018, Lusso et.al 2019). In this work, we demonstrate that the log polynomial expansion generically fails to recover flat $\Lambda$CDM beyond $z \sim 2$, thus undermining the $\sim 4 \, \sigma$ tension claim. Moreover, through direct fits of both the flat $\Lambda$CDM and the log polynomial model to the SNe+QSO+GRB dataset, we confirm that the flat $\Lambda$CDM model is preferred. Ultimately, we trace the tension to the QSO data and show that a best-fit of the flat $\Lambda$CDM model to the QSO data leads to a flat $\Lambda$CDM Universe with no dark energy within $1 \, \sigma$. This marks an irreconcilable tension between the Risaliti-Lusso QSOs and flat $\Lambda$CDM.
Comment: v3 22 pages, 10 figures, a major revision; v4 wording improved, typos corrected, conclusions unchanged; v5 and v6 match published version