That’s not how it works, especially for Rubin. In practice what it means is much lower statistics on streaky data over time. The streaks are not point sources either, so they have a disproportionate impact. You can deal with this through survey strategy to some extent.
You can almost think of this as something similar to vignetting in the final data products. Certain areas will have lower statistics and especially lower temporal resolution based on the season depending on where they are relative to the horizon near twilight.
So, a single image could be lost, but there is supposed to be 1000 or so good images of that area over the survey, about 100 a year. With the satellites, potentially you’ve now lost 3–8 images a year for any given section of the sky (probably more near the equatorial plane), lowering your statistics of the entire survey 1-10%, depending on the declination.
I’m spitballing numbers, there are actual papers you could read though.
Rubin is “wider, deeper, faster”. This reduces all of those dimensions to some extent, but especially the deeper.
