Honestly, I have no idea. But in my opinion...

Longer term if Microsoft / Amazon refine their offering, nobody else stands a chance at any big growth here. Microsoft have the biggest advantage IMO because they can do cheaper software, they get an amazing deal on hardware, they know the software better than anyone else, they can market it better than anyone else. Marketing is unique because they can change their licensing for non cloud customers to tease then into the cloud, then IT people will want to take the cloud exams to get qualified, which is powerful because when people get qualifications in an area, they tend to become evangelists. Why? Because firstly there's a feel good factor of having the qualification in something that people around you are enquiring about. Secondly they know the product so they're comfortable using it, so they look like better employees. There's risk in putting in another product and looking like an idiot. Thirdly they are forced to learn all the features of the product, so they know the benefits of the product without knowing the technical benefits of the other product (a problem because adverts are salesy and non technical, when the advisor to the buyer or buyer themselves is technical). The exam training is practically selling the products and forcing prospective customers to repeat the benefits over and over in their head to exam cram. Also it makes sense for managers to use technology their staff are trained in.

So Microsoft have the edge in my opinion...

However, at the moment, if their sales are good and they have loyal customers and the purchase is earnings accreditive, or if they can offer something the big guys can't (local datacenter, access to said datacenter, support, extra services such as engineer resources for projects, data sovereignty, etc) then they may be sorted. Also they may have barriers to exit for customers, such as difficulty migrating out of the datacentre due to size of data stored and inability to migrate it out without an outage, fear of change. There is a comfort factor of knowing where your datacenter actually is and knowing the people hosting it.

Also Microsoft might not sort their $#!& out and might not engineer, maintain or push the product properly. Nothing is certain, right? It did take an unusually large amount of time to get azure out of beta.

All that being said, I would guess that there will be a place for small, local datacentres such as the one IFT have just bought. It's just not likely to be the next XRO.