No. When so called small sample sets are portrayed as “data” and they lead to conclusions that are not logically consistent with observed facts, the whole think has the whiff of month old milk.
Fact
Trump has not won one state by more than 50%.
Fact
Trump has relied on getting around 35% to win a divided primary
Theory
Polls show Trump doing poorly amongst women
Conclusion
If the polls ate correct, and women hate Trump, he will not be the nominee