reviewers parser should ignore wrong cases

[beliaev-maksim]

We have an issue where some people are using the Reviewers metadata incorrectly. For example, putting multiple reviewers in one cell instead of one row per reviewer.

Basically, if the entry is not a single email, it is incorrect.

We need to fix this, and the approach is still to be decided. Options:

1. Ignore incorrect reviewer entries in the parser
2. Ignore them in the parser and leave a comment on the doc
3. Ignore them in the parser and, using suggestion mode, delete the line in the doc with a comment