The commitment clearly states in the preamble: “to every guideline there can be justified exceptions”. If you can convincingly explain why your data cannot be published due to anonymity reasons, it is perfectly OK not to comply to point 1. If you sign the commitment, you simply should give that justification in the author note, or in the README file of your public project repository.
You should also think about making aggregate level data open. For example, if you analyzed your interviews by counting certain topics or keywords, you could probably publish these summary data without compromising the anonymity of your participants. This way, at least your quantitative analyses can be reproduced.
In any way, this would not be a reason not to sign the commitment.