Having looked pretty deeply into Vit D studies a while back, my takeaway was (in broad strokes):

* Adding more Vit D doesn't appear to meaningfully prevent upper respiratory infections.

* However, being deficient in Vit D appears to be correlated with a meaningful number of D deficient people who do happen to get an upper respiratory infection having a more serious impact. Basically, a higher percentage of deficient people (but certainly far from all) appear to have worse symptoms for longer and have more trouble fighting the infection off. Whether that just means an extra day of feeling under the weather or that you join the very small percentage of those hospitalized for upper respiratory infection depends on the other factors you already know (age, overall health, etc) much more than on Vit D deficiency. But Vit D deficiency is very probably in the top 5 somewhere behind the two statistical Godzillas at the top, age and overall health (which are overwhelmingly correlated).

This is only worth talking about because a fairly significant number of people are Vit D deficient at least some of the time (probably more than 20% but less than half). Who and how much depends on where they live, lifestyle, age, diet, time of year and there is also a genetic propensity for deficiency that's primarily based on race.

Bottom line: If you're Vit D deficient then it may be a good idea to supplement a little Vit D which is cheap, easy and extremely low risk (Note: nothing is ever zero risk across a large enough population but supplementing a small amount of Vit D is about as safe as these things ever get). It may be especially worth considering if you're in a statistically higher risk group, location and/or season. So, if you're at elevated risk and can't be younger or healthier, at least don't be Vit D deficient too. Once you're not deficient, taking even more Vit D won't help more (and can be harmful). Mega-dosing can definitely be harmful, so please don't. Starting oral supplements once you have symptoms is also too late to matter.

In terms of confidence levels, my sense was there's pretty clearly a meaningful correlation here but causation and relative effect are fuzzy. The correlation mostly comes from looking retrospectively at the Vit D levels of those who have infections serious enough to get hospitalized. However, the Vit D correlation is far less than age or overall health (which are overwhelmingly large). The challenge is looking at it the other way, from the front end, where it gets pretty fuzzy trying to tease out high confidence causal data, narrow other factor's contributions or derive a degree of impact specific to Vit D. There are a lot of potentially confounding factors and Vit D is not usually checked in many blood work panels unless there's a reason to. Worse, long-term diary studies of diet are notoriously inexact. So much so that as an armchair amateur scientist just trying to objectively assess data, I have to attach error bars to longitudinal diary diet studies so large they usually swamp any signal. For example, the question: "Does being Vit D deficient meaningfully increase susceptibility to infection?" I don't think there's sufficiently clear data to make any judgement.