When you rent each carriage from a rolling stock company you’re incentivized to run as short trains as possible to keep them as full as a sardine can. Renationalize the whole lot.

Longer isn’t always better for trains. Longer trains need bigger platforms and stations are often the most expensive parts of the network. Big trains also weigh more, so they accelerate slower and have longer trip times.

Weight savings can also translate into cheaper bridges and viaducts, though that’s only true for dedicated track. If you’re sharing with freight they are always going to be far heavier.

Much better is upgrading signalling, automating the train, and running far more service because frequency is king for public transit. Many short trains at high headways vs less frequent long trains has the same total capacity but the short, frequent trains provide far more value. When people don’t have to think about the schedule and can just show up and ride public transit is great.

There are also limits to the value of standardizing rolling stock. You don’t want every train set to be bespoke, but there is also danger in every train being exactly the same. If there is a parts shortage or identified design flaw your whole fleet can be grounded all at once. If you have some diversity you can limp along if one design has to be sidelined.

At some point the train is longer than the station platform. Then you need to extend the platform, or passengers have to know that they need to be in specific cars to be able to exit.