> So, just because it’s not perfect we don’t do it at all?
Just because is so far from perfection to a point of being just waste of bandwidth resources, a security risk [1] and an annoyance who force people a certain dress code, others with shows of decorated background and animations, others shows misery of their homes while some show their wealth status as a way to impress etc.
> telling people that it will never be a replacement for physical presence, would actually quite support the crowd asking for a 100% return to office fearing that innovation and team cohesion will suffer
Honestly, without much data but hearing climate in my circles and reading the press, those who want back in office are just middle managers and some old people who will retire in few years. So the crowd will be a very small one against a far bigger one who just say, as happen right now in various part of the world "if you want me back in the office I resign, go look for someone else"... You can't made soft paradigm changes, you can only change softly witch means for instance start pushing 99% remote work just to a series of jobs where people have enough money, interests and willingness to live cities and going full remote instead of pushing it worldwide "for the pandemic" and for all possibly remote jobs. For instance instead of pushing crappy proprietary platforms like Meet, Teams, Zoom developing a company infra on company iron, negotiating seriously with workers: "if you want full remote you need to offer us a spare room in your house at a given little monthly fee, like the same you get in office to eat, with possible sufficient connectivity to work, electricity and a closed door. We gives you all the needed furniture and gears, we pay the connection or we negotiate a compensation for them if you provide them and what you offer meet our demand" instead of giving craptops and craphones etc. When a first batch of humans, skilled, educated and wealthy enough to test the new model will finally arrive to a good stability than offering a similar paths to other workers also a bit at a time to make the migration possible, the softer possible, with enough experience already made to avoid making full-society-scale or full-company-scale disasters. You can't switch paradigm in small steps though, hybrid work does not work, mixed work in the same workforce does not work. It's a classic error common in countless situations, from the military where a general want a strategy another an almost totally different one and the chief finally choose a middle ground between the two making both a failure to the classic IT horror stories and anything in between.
Consider a thing: there is no success without risk, you can mitigate the risk reducing the risk surface, like "hey we build masonry houses only, we want to try new materials like wood, so instead of change all houses a big risk surface, we build just some few houses to see how they work" but you can't get anything from "we start with a wood based floor and one masonry".
[1] you know deepfakes need much data to be trained, the more you face/body is shown on-line the more data some model get to produce a realistic video of you stating terrorist intention or perhaps confessing a love betrayal etc realistic enough to be believed. In real life no one can hold good cam and grab data without you noticing, with modern platforms like Zoom, Teams, Meet, their owners harvest gazillions of data, and deepfakes aside what you think about holding a talk in a public place discussing works stuff? You normally want a private room, isn't it?