Exactly. Neither should aggregate numbers that various US states and Prof. Spector use to spread fear, acrimony and stigma should be used to estimate VE, or mislead the audiences to estimate VE by themselves. Notice how the ZOE video you linked lacks the intellectual honesty to display the same disclaimer. His own website says "Incidence estimates were updated on 21 July 2021.". He's basically fabricating numbers, for all we know just as accurate as Prof. Ferguson. This is no better than the "antivax garbage" some people are so keen on summarily calling out.
Here are some longish term studies and preprints for Pfizer's VE against infection.
UK, 3 months VEaI = 75%. https://www.nature.com/articles/s41591-021-01548-7
Qatar, 3 months VEaI = 70%, 6 months VEaI = 12.8%, https://www.nejm.org/doi/full/10.1056/NEJMoa2114114
Sweden, <50yo, 2-4 months VEaI = 83%, 4-6 months VEaI = 50%, https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3949410
US, veterans, 3 months VEaI = 78%, 6 months VEaI = 50%, https://www.medrxiv.org/content/10.1101/2021.04.20.21254636v...
Two notes:
* VE against infection drops over time. Sadly, I am not aware of any reliable literature for VE after 9 or 12 months. Unfortunately Pfizer killed their own critical trials, which had a solid headstart, by vaccinating the control arm after 6 months. But the trend is clear, and worrisome.
* 80% VE => 1:5 infection odds. 75% VE => 1:4 infection odds. 50% VE => 1:2 infection odds. 12% VE, it's a wash. Numbers like 1:6.8 infection odds this long after the vaccination campaign are fantasy.