I’m not convinced it’s close, at least in terms of the popular vote. Those polls are well dodgy.
1. Many of them are based on likely to vote respondents, which excludes those who did not vote last time. So all those young people and suburban women who have recently registered are not in the sample.
2. Many polls still rely on landlines. Young people tend not to answer unknown callers whether landline or on-line.
3. Most polls make adjustments which assume they are under counting Republicans, this can be totally wrong.
4. Many polls are paid for by parties which throws doubt on the results published.
5. There is the phenomenon of polls tending towards the same results.
6. I suspect many respondents are hiding their true allegence from their Republican family.
7. Etc etc.
On the ground, the evidence points to Harris: enthusiasm, rallies, local active offices etc.
Vote blue.