One significant difference in this study is that it focused on people with low baseline vitamin D (10-30 ng/mL 25(OH)D), and moderate intervention (2000 IU daily).

The meta-analysis you posted did perform subgroup analysis on people with low baseline vitamin D (<25 ng/mL), but this included a wide range of intervention levels, 90% of which were <2000 IU daily equivalent. They also performed subgroup analysis on high intervention levels, but this included a wide range of baseline vitamin D, 90% of which were >25 ng/mL.

And also there's a difference between infection incidence, intensity, and duration, the evidence I've seen has been strongest in reducing intensity and duration. Also dosage might just be too low.

I've been feeling a little off lately with some respiratory symptoms and took 25,000 IU of Vitamin D, in people who are deficient (probably me lately) 400-1000 daily dose might not actually do enough to have an effect.

It's about time for a meta-meta analysis comparing the traits of the different sets of papers (N, dosage, deficiency status, time of year, duration/incidence/intensity, etc)