I am not sold on full remote working for most companies. I am not even convinced the little covid experiment demonstrated anything, as companies still benefit from employees knowing each others from before the lockdown. Long term the effects in term of productivity, employee morale or culture may be quite different. A compromise is having more local offices, but even then, everywhere I saw it before, it creates a “them” vs “us” between offices, people feel remote from the centres of decision, etc.
But -- and I feel like I'm alone in this -- my personal observation is that even for teams that knew each other before full time WFH I've noticed the teams _mostly_ are not doing new and challenging projects. There's a ton of work and small iterating possible in any company, but I notice all the challenging projects have been punted - or are moving very slowly.
My personal conclusion is that remote work does not, in fact, work well, at least for some types of projects. I can totally see UI-heavy stuff working out, for example, but not embedded or high complexity projects.
A high local real estate price is a reason you can't profitably work for less and its something you can profitably express to an employer if they will take it into account in order to access the local labor market.
A lower local real estate cost is a reason you could work for less not a reason you ought to want to.
You absolutely should try to have it both ways to the greatest extent possible because they are too.