This problem doesn't apply to just break points on return, but also on serve, meaning Djokovic is neither saving nor converting break points at a high rate.
In his career, Djokovic saves about 65% of all break points he faces, which is almost equal to his rate of winning non-break points on serve (67%). However, in his last seven matches, Djokovic has only saved 52.5% of all break points he has faced, which is drastically lower than how well he is doing on non-break points.
Djokovic is facing a similar problem with his return. In his career, he has converted 45% of all break points he has faced and in 2011, he won 49% of break points on hard courts. Even in 2009, which was Djokovic's worst year, he was winning 42% of break points. Yet since his match with Nadal at Rome, Djokovic has only converted 36% of break points.
Earlier this year, Djokovic won 17 of all 21 break points against Dominic Thiem to win 6-3, 6-4 despite being thoroughly outplayed on non-break points. There is a chance Thiem will play Djokovic in the semifinals and if Djokovic does not raise his level on break points, he will have no chance against the Austrian.
The way the Serb has been able to survive so far is that he is producing far more break opportunities on his return than he is facing on his serve. However, continuing to do that will become significantly more difficult in the later rounds of Roland Garros as the level of his opponents rises. In order for Djokovic to win the only major that he is lacking is to improve his level of play on the break points, which seems to be the only hole in his game of late.