"With infinite money, would you rather create a team that works in-person or remote-only?". I bootstrapped my SaaS to $600k ARR with a remote team before switching to "in-person core + satellite", and I'd never start with remote-first because how much harder it makes
everything (from comms, to on-boarding, to training, to culture, to payroll, to knowledge transfer, to customer support etc).
My position is: the founding/core team is more effective (i.e. = motivated, engaged, and better aligned) when they work in-person at least 50% of the time. Two days a week, or more. I see this as axiomatic, because we are social creatures and the nature of innovation/productivity requires rapid exchange of ideas among developers/sales/marketing/founder roles. If we do not agree on this, then the rest is irrelevant.
If we do agree on that, then the natural conclusion is: if money is no object, you must structure a startup to prioritize in-person collab (but offer remote days, because psychology is hard). In the real world, money is an object most of the time, so this is when remote-only discussion start ("May be if we talk on the phone every morning..." and "Do we need to fly in 4 developers for a month, or can the screen share?" etc). In the VC territory, money is free and time is priceless. It makes sense to spend extra ~few hundred thousand/pa to have the team co-located if it increases odds of success by >5%. That's it: the essence of my cynical argument "this is purely about costs" actually means "With infinite money, one strategy is more likely to deliver superior results. Now we guess how much more likely, but FYI serial founders and people with ~infinite capital almost universally choose the in-person strategy even today".
Lastly, an appeal to authority. Pick someone you respect, and see how they actually work. They likely avoid meetings, and sing the praises of focused/remote thinking, while going to the office every day to sit down face-to-face with the people that move the lever the most in their business. For a small startup, that's your first 5-10 people. Try to see the whites of their eyes as often as possible.