The author didn't say that "performance didn't matter" as a starting assumption. He said it as a tentative conclusion, that specifically a small improvement in processor performance was not the most important factor in the user experience, or in adoption by tablet makers.
How could performance be unimportant if all of the high ASP tablets are using performance as their key differentiator? Obviously, lower cost tablets with less performance are growing at an incredible rate - but it appears to not have been at the expense of high end tablet sales (at least, not yet). It's also interesting that Microsoft has gained 6% of the tablet market (in the first quarter of high volume ramp, by the way) with Windows 8 based on a system price that's much greater than the average Android tablet, with Core based tablets (such as the Surface Pro and Samsung/Acer Ivy Bridge tablets) selling strictly on behalf of performance, at incredibly high price points, and driving some of that gain.
Seems like there's already plenty of data to suggest that part of the market buys on behalf of performance. Not all, but we should expect that segmentation will continue in the future, and not all race to the bottom. I believe the data will be even more obvious once Bay Trail and Haswell tablets start to ramp in the market.
Market Data 
Markets 
AI
