Safety in numbers?

I see the point of this bit from John Gruber (quoting Harry McCracken) about iOS vs Android and what the numbers say:

Great work, and his conclusion seems perfect:

Android if you’re talking about market share; iOS if you mean financial success. So far, this is a strikingly different market than the PC business back in the 1990s, when market share translated directly into financial success.

This is true. So far. But consider the IT hardware/software business back in the 90s:

In the late 80s and all through most of the 90s, the Unix vendors (Sun, in particular, but also HP, SGI, others) boasted margins in the server market that underwrote their intentional shunning of the brutal Wintel marketplace and the pennies it yielded (to hardware vendors). In the mid-90s, AT&T was putting Sun workstations that cost upward of $10,000 on the desks of employees and spending a multiple of that amount in the server room for the same hardware. What sane CEO would leave this paradise for price warfare on PCs?

Today, Unix has won (in the form of GNU/Linux and, ironically, OS X/iOS/Android) but the Unix vendors with those fat margins are near gone. They are gone because computing was becoming commonplace and processor performance was sufficiently advanced that Wintel’s price advantage and user friendliness made it ubiquitous.

I am not implying that this holds a clear message for Apple, that they should, as the pundits have been baying for, introduce low cost models of iPhones and Macs to gain marketshare. Rather, I think that simply as a matter of correlation, the Apple vs Android situation today is not dissimilar from Unix vendors vs Wintel in the 90s.