Welcome to the forum.
Re: “I think it should be rectified or mitigated for the upcoming 2023 standard,” per the “rules” it is not possible, moreover there is no guarantee anything will make it into Fortran 202Y revision and then whether and when compiler implementations will fix the processors, so the poor, persevering practitioners of Fortran may be glaring at the “hole” you describe indefinitely.
As things stand, the revisions to the Fortran standard are on a 5-year cycle of which a considerable time is taken up by various administrative aspects, perhaps even upwards of 2 years’ worth of duration. Since late last year 2021, clearly early this year 2022, the Fortran 202X = 2023 standard arrived at a “code freeze” stage, meaning only wording changes, etc. and minor editorial aspects only can be made.
Moreover in the certain areas like the so-called /DATA subgroup, the one that would ordinarily address the kind of changes to the PROTECTED semantics you highlight, the real work around papers and discussion and votes can only be truly achieved when the chairperson of the subgroup is available which is mostly 4 actual, physical, face-to-face meetings a year over about 3 years of development duration (in a 5 year-cycle) with roughly 4 days per meeting, give or take about 400 hours of time and attention. The same person has multiple responsibilities and there are various other matters to address also. This is yet another resource constraint. Thus, like the wheels of justice, things churn slowly and get delayed often just like justice, and which means the practitioners get denied.