It's unlikely that people will reject centralized power due to "accidents". The issue is that there are significant benefits to scaling up your power station no matter what technology you use, whether that's combustion, hydroelectricity, or even solar or wind power (not that solar is particularly practical at this point in time). Larger and more remote power stations are usually more efficient than smaller and more decentralized stations through a combination of factors: larger furnaces and boilers can produce more heat relative to their heat loss rates, they require relatively less power conversion equipment to hook into the grid, they can be located on cheaper land in rural areas, relatively less labor is needed to build them (the striking cost difference between distributed and utility-scale solar is heavily driven by labor costs). In the case of hydroelectricity specifically, there are geographical conditions that mean that certain areas can produce enormous amounts of power, and you would be foolish to reject those in favor of much smaller plants if you can economically build the full-scale one.
Of course, these benefits don't go up forever, and building ever-larger creates its own problems, but it's telling that most power plants are in the size of a few hundred megawatts to a few gigawatts in scale. That seems to be about the most economical size for a power plant, not so large in most cases that it becomes a single point of failure, but not so small that it multiplies expenses unnecessarily. Of course you will have exceptions in both directions, particularly for hydro, where there are a lot of rather small systems out and about and of course gargantuan plants like Three Gorges or Itiaipu on the other end of the scale. But by and large "moderate" scale in a relatively centralized system will probably rule the day.
