I think there is a downside.
There needs to be a method by which the ACM, or any other group, can protect its name and reputation should a member or group of members do something abhorrent. For example, computing professionals who, in their line of work, design, build, deploy, and maintain software designed exclusively to violate the law without speaking up to their superiors, to any applicable regulators, or to the general public (likely through the press). If it became known that someone did this and was a member of an organization, it could paint the organization in a bad light, affecting the reputation of current members and driving away potential future members. Having a documented set of standards would allow the ACM to disavow these "bad apples" and terminate the relationship.
I think that there needs to be some minimal standard that is required for ACM members. Although I don't disagree with anything in the current code of ethics, I do think that it is far more reaching than is actually needs to be. This proposed draft is not only far reaching, but has ambiguities and controversial statements that make it difficult or impossible to enforce. These controversial statements or overreaching requirements are the kinds of things that I would expect to either be ignored or to drive members away.
I'm still looking at the IEEE Code of Ethics with its 10 statements, the British Computer Society's Code of Conduct with its 4 principles and 22 total statements below them or the Australian Computer Society's Code of Professional Conduct that have one sentence high level summaries and a long form with between 4 and 9 detailed requirements. I think these are more of a model of what I would expect for a code of ethics that an organization would require of members.
When I look for aspirational ethics, I look toward the Software Engineering Code of Ethics and Professional Practice. It's much longer and more detailed than every organization's ethical code that is required of members (except the ACM's, of course), and isn't tied to membership in an organization. This is the kind of opt-in code of ethics that I would expect. It's accessible and can be promoted beyond a single organization as a guide to professional conduct.
I agree with this. I am opposed to mandatory licensure for all software engineers and I think that organizations having and enforcing codes of ethics can reduce, if not eliminate, the need for licensure. Although I also think that would require that both academia and industry recognize the value of membership in organizations such as the ACM. If organizations had reasonable, enforceable codes of ethics, then being a member in good standing of those organizations should carry some weight with employers and clients.
At the same time, I also see a need for an even higher level of ethical behavior. Like I mentioned before, the code of ethics for members is, I think, to protect the organization's reputation first. Having a recognized code of ethics that is more aspirational or visionary can be useful, especially if professional organizations not only identify the bad apples, but identify and honor the people who make the good ethical decisions. Again, I'm looking to the SE Code for that kind of visionary ethics.
The SE Code was last updated in 1999. I would prefer if the ACM once again collaborated with the IEEE Computer Society on an update for it, but I also wouldn't be opposed to the ACM going forth on their own and coming up with their own aspirational or visionary ethical code. But that code should be separate from the basic requirements of membership in the ACM and be more voluntary and opt-in.