Well, there are two major challenges to get this approved. First, on the technical front, despite all the hype about CRISPR, its off-target editing effect is real, and it would be too high to tolerate in germline editing setting. That being said, Chinese research was the first to edit human embryo a few years back, and two Nature papers last year showed some gene editing in human embryos. Still, the tools need to be improved. Second, on the ethical level, it would be extremely hard to get anything involving human embryo to be studied in clinical setting, not to mention gene-edited embryo, in western countries. All said, technical difficulty is still there but can be solved sooner or later. Ethical difficulty? Not in the US (or Europe) in foreseeable future, but Asian countries would be a different story.

It's already happening in China.

China was the first in the world to publish human germ line genome editing using CRISPR/Cas9. In 2015, a group in China reported editing in defective human embryos. In 2016, another group in China reported editing in viable human embryos. These are scientific articles by academic labs; one doesn't know what's going on in other research sectors. Some clinicians have recently been found to have performed mitochondrial replacement therapy before it is fully developed and approved, the same is most likely true of CRISPR germline editing for human single gene diseases. CRISPR has been used to edit blood cells in China since 2015 as part of a cancer clinical trial. It is reported that it took one afternoon to approve it (vs. two years for something similar in the US). So I think the answer to your question is: soon!

With mitochondrial replacement therapy one is already permanently changing the genome of the future baby. Just not the nuclear genome, but the mitochondrial genome, yet it also codes for genes and can carry disease susceptibility variants... If both mitochondrial and nuclear germ line changes are done with the purpose of correcting genetic disease, how different do you think the regulatory milestones of approving either will be?

Permitting mitochondrial DNA edition is a lower bar, because
1) male babies will not pass their new mitochondrial DNA to the next generation (Germline transmission), only females
2) mtDNA only contains 13 protein coding genes, and all these proteins are part of OXPHOS. But still, why is modifying those genes any ethically different than modifying genes involved in other cellular pathways? Given that the goal is to prevent disease, not engineer specific physical traits...

As a lay person following the development, and use of CRISPR technology and methods, I believe the question is to the subject matter experts, when can we expect sustained viable solutions that for even the simplest defect corrections to be medically acceptable treatments on a large enough scale such that they will achieve routine use in patient treatment plans. What is the timeline looking like?