I'd say you are wrong that they had a long run without serious competition. They had serious competition from day one: Other hosting businesses ranging from colo providers to managed hosting providers. They made a splash by carving out a new niche. Some people have done well entering that niche (so far), though many of them were well capitalized existing players taking advantage of their position (existing hosting providers adding cloud features to take advantage of existing as-yet unsold/unrented stock to make it cheap to enter this market for example). But the number of new entrants in this market since Amazon is still vanishingly small compared to the number of people salivating over the smartphone market.

But Amazons strategy still puts them in a situation where going head to head with Amazon in this space now puts you in the situation where if you can't compete on features, they will keep chipping away at your ability to extract a high margin.

At the same time they've gradually carved out a larger and larger niche: Additional products, and ways to improve pricing such as having people pay for reserved instances.

This is similar to the physical object space - there were lots of people that could compete with them early on, when Amazon was still mostly selling books, and selling small enough amounts of books that they could not push the distributors around.

But even then Amazons model is devastating to competitors: If your cost base is lower than Amazons, but your investors have come to expect a 5% return every year, yet Amazon gets away with staying around 0% without getting any flak, you are screwed unless you seriously believe that you can continue to keep your costs sufficiently below Amazon to be able to compete on price and retain your margins.

One by one Amazon competitors have fallen because it is incredibly hard to reduce bloat once it has become part of the way you operate, and incredibly hard to wean yourself off higher margins. When they then go up against an organization whose credo is based on cutting cost everywhere it is possible, and then extracting almost nothing in return for it, a lot of people will be in deep trouble.

I don't think Amazon will ever be as devastating in the hosting space as elsewhere, as there are too many ways to differentiate in the hosting space. But they certainly can keep chipping away at the core, and if I was working in hosting, I'd be spending a lot of my time figuring out 1) how to cut costs, 2) what product categories Amazon are unlikely to want to be in soon and/or which doesn't fit with well with Amazon's model (for example anything where customers wants to pay for a lot of reassuring face time and handholding)