There are few who can (arguably) create value by combining ... candies and railroads like Buffet.

I would assert that it's not really that difficult at all. Doing it that outstandingly is impossible, but doing it materially well is not a big deal.

The big advantage is to have a culture of rational capital allocation. Don't give more capital to a unit that isn't investing it with decent returns, strip them of their FCF and send it tax free to head office, which will invest in the best available opportunities. This is the one big advantage for any conglomerate.

The main thing is that the vast majority of firms are not rational capital allocators, they are empire builders, allocating capital to things that are either "more of a bad business" or fads or vanity projects. Consequently attempts to garner the conglomerate advantage generally fail for them.

Given the weak competition in this regard, and Berkshire's very deeply seated tradition of rational capital allocation (and even acknowledging that it's a thing!), there can be a pretty decent advantage for a long time post Buffett. Don't let a unit reinvest badly, and don't let head office get sucked into bubbly optimism when picking targets.

Jim