This happened in 2013 in the context of finding $\epsilon$-approximate solutions to the maximum $s-t$ flow problem in near-linear-time (taking day to mean any 24 hour period).

The two references are https://arxiv.org/abs/1304.2338 (Jonathan A. Kelner, Yin Tat Lee, Lorenzo Orecchia, Aaron Sidford) and https://arxiv.org/abs/1304.2077 (Jonah Sherman).

As usual, the timing was not a coincidence.