It looks like the structural advantage for the GOP from the electoral college is continuing to grow. The structural advantage can be defined as the difference between the popular vote for the country and the vote in the tipping point state. In the last election it was 3%. Pennsylvania was the tipping point state and there was about 3% difference between the national vote and the vote there.
In this election Pennsylvania still looks to be the tipping point state. But the structural advantage (or disadvantage depending on your perspective) is about 4% (9% deficit in the national polls vs 5% deficit in PA).
The structural advantage is related to surplus or wasted votes in non-competitive states. For example, California went from being 19% more Democratic than the rest of the country in 2012 to being 28% more Democratic in 2016. That might help in down ballot races, but from the perspective of the presidential race that's a whole lot of wasted extra votes. And the number of these surplus votes grew a lot in 2016 and that contributed to the structural advantage in the electoral college widening.
In 2020, the widening comes from a different direction. Fewer wasted or surplus votes from deep red states, in particular Texas. It appears that Texas is going to go from being 21% more red than the rest of the country in 2016 to about 10% more red in this election. That's a big reduction in wasted or surplus votes from the GOP account. So it pushes the structural advantage in the electoral college further in the GOP direction.