Ok, should we go with your personal opinion then? What percentage of that 76% of votes would you say is accurate? It doesn't matter anyway because the only ones who know for sure are reddit insiders and they won't release that info, they only break it down by app and website not 3rd party vs official.
Bias isn't a personal opinion. It is more likely that a person sees this poll coming from lemmy than from reddit, and people coming from lemmy are more likely to have left reddit due to using a 3rd party app that will shut down. On the same note, someone who is on reddit still is less likely to have seen this poll, and if they're still on reddit, are more likely to not be using a third party app and thus less likely to care.
This is heavily skewed towards more technical users who are more likely to care and use a 3rd party app
Your response reminds me of the people on reddit during early covid. People commonly were using metrics for mortality like # of deaths/# recovered or # dead / total cases and then refused to think otherwise, perferring a invalid metric because the numbers early on fit what they expected rather than realize they may be wrong.