> Also fast cars are a problem when you need to make a pedestrian crossing

The three things I quoted in the post you're replying to were "better pedestrian infrastructure that's actually separate and protected from the road", "Separate that traffic, get the cars out of the way from pedestrian streets", and "infrastructure separating transport modals".

You don't need a pedestrian crossing if you have separated infrastructure. For instance, interstates don't have pedestrian crossings. (Some have raised paths where pedestrians can walk from one side to the other without intersecting with traffic.)

> To get fast from point a to point b you need public transport

That's a lot less fast when the path from A to B involves walking to C, taking transport to D, walking to E, taking transport to F, then walking to B, and taking twice as much time doing so. Even if transportation were instantly available with no waiting when you arrive at each of those points, that's still substantially more inconvenient. And it's a largely fundamental property of public transport that getting from an arbitrary point to an arbitrary point typically involves multiple transits plus walking. (And unfortunately, often the responses to that are some flavor of "we should make cars slower and less convenient" rather than "we should make public transport faster and more convenient and point-to-point".)

It's hard to beat direct door-to-door transportation. It's possible, and we can and should get to a point of having that via public transport, but in the meantime let's not pretend that it's always a win rather than a tradeoff.

Separated crosswalks aka raised paths inside cities are terrible for pedestrians, that's why many cities in eu are either closing them or doubling them with classic crosswalk and finding out that raised/under paths aren't used anymore since it's much more convenient to just directly cross the road

Again, properly designed public transport is faster than cars. You are thinking about public transport in current car designed setting. Imagine each bus/tram has own lane and semaphore priority meaning it'll get close to max speed, imagine thereare lot's of pub transports, imagine the paths for pub transport are shorter compared to car paths again to make pub transport more efficient, imagine parking is limited since land is expensive and youll spend lot of time searching for a spot and it wouldn't be cheap since again land is expensive, imagine in either situation you'll end up spending time in traffic, imagine most of pub transport stations would have bike parking so that you could cover last mile on a bike really fast if you need it

You can say that it'll cost a lot of money and time to implement this but in reality it's just a matter of political will. Separate bus lanes and priority semaphores and bike lanes and parking is relatively cheap and easy to implement, just like dynamic parking price. The most expensive part is buying more pub transport units.

> Separated crosswalks aka raised paths inside cities are terrible for pedestrians

I agree, which is why I personally prefer the solution of burying the roads and keeping the pedestrian walkways at what is currently "street" level. That's a major challenge for existing infrastructure, but I've seen more than a few public transportation proposals that have similar "much easier when one from scratch" problems, and I think it's worth designing the ideal before settling for something worse.

> Again, properly designed public transport is faster than cars. You are thinking about public transport in current car designed setting.

No, I'm thinking about ideal public transport versus ideal car transportation. It's not reasonable to compare the best case of public transport to deliberately worsened car transportation and declare public transport the victor. I would love to have public transport that's actually better than the common case of car transportation, but proposals like what you're describing don't go far enough to get there.

I would love to have a world where we have 300km/s trains between every city (major or minor), and automated point-to-point no-transfer underground transportation within cities. And I'd love to see incremental steps in a direction like that.

What I don't want to see is "if we make cars much worse, we can have public transit that sucks less but is still worse than cars used to be".

Best case of car transport would be if few ppl use it which is achieved by giving priority to public transport and bike paths. If this (car) mode is prioritized, car transport by definition will be a worse experience than an ideal bus because you don't get traffic with the bus. Burying cars under is a good idea in theory but not that great in practice. It's extremely expensive to do it(and also build all the underground destination infra) and in the end you still will end up with traffic, the difference being that all the drivers will be trapped with their fumes/microplastic tire wear underground.

You don't need to make cars much worse, just make pub transport and bike/pedestrian infra as good as possible and give what's left to cars