Serf's Corner 1: Whither OEM? MicroHard?

I'm going to work on a fortnightly open discussion to spice things up. We did it previously and worked well, so here's the first instalment. I'll end each post with two questions for the commentariat.

Along with recent news and Silellak's post, I got to thinking about the classic OEM/Vendor model and about MSFT's future.

History to repeat itself... whither OEM?

Early news is that Samsung's earning are not going to be that good. The reason? Price pressure from the bottom, and struggles in meaningful differentiation on the high end (to compete with Apple). People predicted this. Samsung's margins may well erode over time descending towards the single digit margins of PC OEMs of the past. The end-game is that Android will rule, with OEMs competing on good enough-ness and affordability rather than quality or value. Like the PC market.

Marketshare is a tent, profitability the tentpoles. Is this the time for MicroHard?

For a software vendor, in what world is competing in a market where there is no value attached software attractive? Answer: none. Yet this is the situation in the rapidly growing, and dominant mobile market. Microsoft are trying, as a company traditionally reliant upon huge software margins and with little in the way of the advertising monetisation competencies that Google has. Cheap Windows Phones? So what. Nokia X2? Barely profitable. The current billion are using Android, and the next billion are not paying for Skype, OneDrive or Office 365. In fact, their primary devices are cellphones - where the latter (and most crucial), is free.

The two Qs:

1) on OEMs - if software differentiation is out of the question (i.e. this excludes Apple), how can an OEM differentiate on the hardware side to ensure margins remain healthy? Has this ever happened? Can this ever happen?

2) on Microsoft - the major soft' (Windows) may end up becoming a cost centre rather than a revenue source. Is now the time to utilise the vertical stack for profits? How could this be done?