One definitely does not need to draw conclusions from the percentage of Democrats, Republicans, and Independents polled. Party ID is a dynamic characteristic and should not be weighted for, which is why the best pollsters do not weight for them.
People do this every year, whoever is losing in the polls. In 2012 it was Dean Chambers. "Look how many more Democrats than Republicans are being polled. When I unskew them, (he created unskewedpolls.com), it shows that Romney will win in a landslide. Wait and see." The actual result? Obama won, and the polls, while fairly accurate, actually ended up with a slight Republican bias on average.
In 2014 it was all the Democrats. I read the Daily Kos as they have unparalleled coverage of election news, electoral data, and polling, but their analysis is often way off. There were so many articles and comments about how the polls were oversampling Republicans compared to past midterm exit polls and things like that. "Charlie Crist isn't really down by 2% in Florida because in 2010 41% of voters were Democrats. Blah blah blah." The result? Polls were mostly right and actually had a small Democratic bias.
In 2016 it was the Republicans again. And this is the one everyone brings up because of the overwhelming media narrative that Hillary was going to win in a landslide. The polls themselves didn't tell that story. Nationally, the polls were off by 1%. State polls, were generally nailed perfectly outside the upper midwest. It was there that the polls margins missed, although they had tons of undecided voters, didn't poll during the final weekend, and did not overestimate Clinton's actual percentages in those states, they just underestimated Trump's.
In 2018 it was again the Republicans. "Blue wave is bull****. All these polls are oversampling Democrats. 2016 is proof." The results? Polling generally had a small Republican bias, although it was very strong overall. Slight Democratic bias in Florida. Polls that weight by Party ID, like Rasmussen, had horrible performances. Rasmussen had Republicans +1% in their final poll.
Every single time people try and unskew polls they end up being wrong. It is not a good statistic to weight by. The evidence is overwhelming that those who weight by Party ID perform worse than those who don't.