You don't need to survey the population in order for it to be "proper". As long as the sample size is large enough, random, and from the same population, the statistic should be an accurate representation.
This is correct.
We send the in-game surveys to a random stratified sample of the population across all platforms. With specific targeting to cover the bottom, middle, and top level of players in terms of skill.
We use a sampling methodology as we have mutliple surveys running at once and don't want to fatigue players. If surveys occur too often and are seen by the same people over and over then the quality of answers to them can drop as people stop spending time to think about them.