I’ll recheck the link.
Pollsters admit sampling bias is a huge problem, especially in political polls. Being spectacularly wrong, as you say they were, is not evidence they should be relied on. After the 2016 election there was much debate about what went wrong.
You don’t supply any cite for what fixes were made by which polls. The key point you are missing is that we haven’t had a Presidential election since to validate the fixes that were made. You also don’t yet seem to understand the difference between random sampling errors given the assumption we have an unbiased sample and the potential errors introduced by biased sampling.
You also didn’t seem to get the point that Nancy Pelosi is afraid to have a vote to open an impeachment inquiry. You say:
She began that impeachment inquiry last week.
Yes she decreed an inquiry, but she has not scheduled a vote on beginning an inquiry. Why not? Every other impeachment has opened with a formal vote to begin the investigation. Part of the reason Pelosi won’t schedule a vote this time is that support for impeachment is very weak in districts Trump won.