You may not be due for tails on the 10th flip, but you are due for 9 more tails at some point than heads.
Using similar logic, Foles would be expected to perform worse this game than last game as his performance was very close to having a perfect passer rating. We know Foles is not a perfect QB in term's of passer rating, so he is far more likely to revert to normal than he is to do as well or greater this game.
Getting back to the turnover situation, NE averaged 1.3 forced turnovers per game through 14 games. There are two ways to look at it. The first is that they that is there baseline average that they will maintain (in reality, this year was a down year for forcing turnovers for the Pats). If we set the standard that that's what NE will continue to average, then they are "due" to make up 5 turnovers at some point to keep that average. The other way to look at it is that their average should be re-calibrated to a lower level due to having 0 turnovers in 4 games . . . which would be an average of 1 turnover forced per game.
Given that in the prior 6 season NE forced 29, 23, 31, 33, 42, and 36 turnovers (roughly an average of 2 per game) and they have averaged less than that this year, it is not unreasonable to expect them to have a run of games where they force more turnovers. If the longer data set indicates NE averages 2 turnovers forced a game, then they at some point will make up the 8 turnovers they missed in the past 4 games.
As you said, it may not be this particular game, but the expectation is that they will return at some point to their established baseline level and make up the difference. Certainly there is no guarantee that that would happen, but IMO that would be a logical conclusion to make.