While I agree with their position overall, particularly with the ambiguity of "harmfulness", I think this is not the right tack to take here:
> fails to acknowledge how actually harmful it would be for certain groups of the population to have their real life identity associated with their online identity
No nanny-state-style government is going to buy this argument for a moment, because it requires a certain level of distrust of authority that certainly won't be present in the authorities themselves. Even among the population at large, there are undeniable trends in many countries towards authoritarianism on both the left and the right that are ignored or embraced by a majority (or a large minority) of their respective bases of support.
Combined with prevailing trends against web-based anonymity and concerns over "fake news" and other amplification mechanisms for information warfare, I suspect we will see a further bifurcation into regulated, KYC-style services and unregulated, anonymous communities. Even here on HN, posters and commenters are regularly chastised for failing to adequately disclose their identities when they have a vested interest in the topic under discussion.
Rather, the privacy angle which has been so conveniently mooted by Apple lately is (in my opinion) more palatable to the general public. By giving more responsibility to the service providers, the government is also implicitly giving them more power as well (as censors, if nothing else). But in practice, most people have personally observed that when more power/information is provided to private companies, it leads to a slippery slope of monetization, advertising, and lock-in.
An appeal to this sense of unfairness and power imbalance with companies seems more "sellable" to me than convincing people their governments have nefarious/censorious motives. As a supporting argument, consider the general global consensus on gun control -- if people were willing to believe that their governments were ready to turn on them (in terms of physical oppression) at any moment, there would be more global concern about the physical power disparity between police/military and the civilian populace.
Currently (with some exceptions, e.g. the U.S.), the global consensus is that people are much more willing to trust their police/military with weaponry than their fellow citizens, based on the implicit notion that the likelihood of experiencing harm from the latter is disproportionately more likely than the former.
In contrast, as partially evidenced by the relative growth of Signal as a response to the changes in WhatsApp and the public support for app-tracking changes in iOS, people are demonstrably concerned about privacy in the commercial space.
Thus, I theorize that an appeal to privacy would be more effective: "Under the proposed changes, we would have the ability to do ____, which would exploit your personal, private information for profit. If we can do it, that means ____, ____, and ____ can do it too. Do you want your government giving even more power to private companies to profit off you?"