To explain why this isn't necessarily the case in america [0], it's important to understand that america didn't have a huge rebuilding effort post-war [1] that relocated industry away from the rail lines that had been servicing them for a hundred years. Post-war europe did have that rebuilding effort, and it saw a lot of medium industry spread out as the costs to remain in denser populated areas near rail hubs increase. Thus road networks became the feeders for these industries rather than proximity to rail freight depots and rail spurs. Maintaining wagon-load rail delivery in that environment became much more expensive as a result.

The height of the railroad era in the US actually saw the peak of the horse population -- moving goods the final or first mile, or block, to and from the railroad depot was a huge need. The automobile and lorry rapidly picked up this niche, but horses were used well into the 1920s.

A train is an exceedingly good way to get goods or passengers from near point A to near point B. But if you want specific pickup and deivery, you either need to put everthing at ponts A and B (densification, tower construction), build an immense set of spur line (low raffic, high loal impact, low rate of return), develop an exceptonally efficient intermodal system (goods can coninue across multiple transport modes without major break-bulk costs -- shipping containers are an example). Or you have a means of accomplishing the final and initial delivery legs.

Keep in mind that there are transport needs that concentrate poorly, with agriculture being a prime example. You might even call ag an industrial productive mode dominated by spatial (or areal) inputs. Keep in mind that the fundamental unit of area under old units, the acre, is literally defined as a daily work unit. In German, that is literally true: Tagwerk, a day's work.

Which means that both the inputs and outputs of agriculture must be delivered over a sparse network not amenable to infrastructure-based optimisations. And running rail to every last field is a losing proposition. Asphalt or gravel roads are far less expensive infrastructure, and even electrifying those would be a high challenge.

(Other materials sourcing operations face similar dynamics.)

Here's some handwaving: "We have a large existing road network and some heavily used spots. How well would it work to charge trucks while driving along the densest 0.7% of the road network?" Does this sound like a question one might spend money to answer?

One of the big issues is it that is very hard to build new railway tracks. The current German infrastructure is already near the limit and the time and cost till a project starts (a lot of issues with people living near it, costs, regulations) is very high.

All in all I think it's a good idea to improve existing infrastructure but not sure if it will scale.

However what is doing them in is the same which will do in most rail, whether cross country or in city. Efficiency of trucks and buses is increasing and the move towards electrification will remove the pollution then combined with autonomy and specialized vehicles will arise to displace both freight and passenger use

[1] https://www.freightwaves.com/news/railroad/us-and-european-f...

This solution is the result of looking at the situation as it is right now (logistics mainly uses trucks, can we avoid changing that? We don't have land for new railways, can we avoid using extra land?) and trying to make an evolutionary improvement.

Also, because they can take much steeper inclines and tighter corners, trolley buses can compete with trams in busy city centers if those cities have steep hills.

The problem is that railways have massively higher investment costs than roads in construction plus there's roads everywhere but you can't build a freight rail line into the center core of Munich, for example.