How to avoid and address a lack of gender diversity in grant proposals?

This is a hard and important problem. There are a few things you can do to address it.

First, admit that there is a problem indicated by the demographics of your discipline. Tell the funding agency that

neither of the institutions involved have female staff eligible for the grant scheme working in the area of our proposal.

Tell them briefly about your universities' plans for future diverse hiring. Tell them that the grant applicants do not have control over the diversity of current staff (if this is true).

Explain how the project will contribute to gender inclusivity. You said you plan to hire one postdoc. How will you encourage gender minority scientists to apply for that postdoc? How will your selection process prevent discrimination? How will that postdoc receive professional development that will advance their career?

Will you be communicating about your project to the public, and will that be done in a way that reaches people of all genders?

It sounds like your project is applied. Will the application benefit gender minorities? Even if they get a minority of the benefit, it's still worth pointing out that your project does not exclusively benefit one gender.

Do not add a token applicant. You are obligated to allocate the project funds efficiently. Some people argue that tokenism has the opposite of the intended effect.

This can be answered with a pretty simple question:

Are the PIs the best people for the job?

If they are, then, quite frankly, there is nothing that needs "fixing" (at least in terms of this particular proposal, you may have wider problems in your field but they aren't your problem alone to solve and I digress).

If they aren't, then you should look at who is and have them on the project instead.

Whatever you do, don't add a female PI who isn't qualified or suitable for the position just to please the reviewer. It's not helpful and is actually rather patronizing and offensive. Also, dropping a PI just because they're a man and you need a woman is discrimination in and of itself.

Tokenism is just another form of sexism

In the long-term, it might be worth analysing the root causes of why the 5 best people all happen to be men.

However, that's not relevant for this proposal and isn't going to be a short-term fix. I'd suggest you write a letter to the reviewer indicating that you believe the 5 PIs are the best people for the position and leave it at that.

I once submitted a math paper to a top 5 journal, and when the reports came back the referee said that it was more appropriate for the next tier because 1) It only constructs one object which was already predicted to exist and 2) It uses an extensive computer calculation. These were undoubtedly true facts about the paper, and certainly it'd be a stronger paper if it constructed a family of objects or if it didn't require a computer. Nonetheless, it was a very good paper, and there was nothing we could do to address either of these criticisms. We'd thought a lot about whether we could make either of those improvements and we couldn't. 10 years later I still think about whether we can fit that object into a family and I still can't do it, 10 years later I still think about whether there's a less computational approach and there still isn't. Nonetheless, despite there being nothing we could do to address the criticisms, they rejected the paper because top 5 journals have plenty of great papers and can afford to reject your excellent paper because of valid criticisms that you can't do anything about.

All of this is just to say that you should really seriously think about their criticism and see whether you can address it in a productive and substantive way, and if you conclude that you can't do so, then you need to accept that they might reject the proposal simply because they have other proposals that are just as good and which also broaden participation of women in your field. Next time you have an idea for a proposal think about including women co-PIs earlier in the process and maybe you can find a different local maximum (somewhat different subject, different direction of collaboration, different mode of collaboration) which avoids the drawback identified in this proposal. By all means follow Anonymous Physicists excellent suggestions and maybe they will fund the proposal, but in the end it comes down to their judgement and all you're really owed here is that they explain their judgement and that this judgement be based on accurate and valid reasons. In this case they've identified a valid criticism and communicated it to you, and ultimately it's their judgement call.