Right now, we have two indicators: The way the swings on the special elections are going, which are all violently swinging away from the Republicans, even if Republicans end up winning, it's barely by single digits in double digit districts.
And the polling, which in aggregate, across all demographics, has been continually moving more and more negative each passing month. Even polls like Rasmussen, which have a traditional right-wing bias, have shown a consistent drop in support.
Metrics like the ones you mentioned aren't useful as predictors. Yes, the Democrats spent a lot of money. So did the Republicans. Big money elections are the norm now, and each election has a loser.
Your reference to losing special elections only tells half the story. These are races that should have been as safe as safe can be. That's why they're special elections, because when Presidents make appointments, they only tend to choose people from their party who are in such safe districts that they can easily be replaced. Often-times, the national party doesn't even send monetary support to these special elections, because they're re-won so handily.
Take Ossoff's challenge in Georgia. He lost by 4 points in a district that went for Trump by 19 points. Moreover, the Republicans were required to spend $25 million in order to defend a seat that they had most likely budgeted around $0 for. That's what's called a weathervane. Or a canary in a coal mine.
Even further down this rabbit hole, the 0-6 loss article is from June. Since then, we've had Phil Miller win by 9 points in a district that Trump won by 21 points, which means that's a 30 point swing! We also have Kevin Cavanaugh winning by 23.4 points in a district that Clinton only won by 0.1 points. There's also been sweeps of state legislatures in traditionally VERY deep red states such as Oklahoma.
Every single trend points to a very difficult 2018 for the Republicans, and an especially difficult 2020.