They even have a Twitter-vaccine study from *Iran* of all places (!!):

Twitter sentiment analysis from Iran about COVID 19 vaccine https://pubmed.ncbi.nlm.nih.gov/34933273/

"Methods: First, a total of 803278 Persian tweets were retrieved from Twitter, mentioning COVIran Barekat (the homegrown vaccine), Pfizer/BioNTech, AstraZeneca/Oxford, Moderna, and Sinopharm (imported vaccines) between April 1, 2021 and September 30, 2021. Then, we identified sentiments of retrieved tweets using a deep learning sentiment analysis model based on CNN-LSTM architecture. Finally, we investigated Iranian views toward COVID-19-vaccination."

If random twitter is good enough to assess people's sentiments, surveys asking specific questions are surely at least as accurate at conveying people's experience/knowledge