"By making MobileMe free, those using it with iOS devices won’t be using services from Google or Microsoft, which makes switching to Windows Phone 7 or Android more difficult. While PC users would also have MobileMe free, they’d need to have iOS devices to make it really worth using. The Halo Effect, which argues that iOS device sales later lead to Mac sales mitigates the loss associated with giving away MobileMe to PC users in the present. If they do switch, free MobileMe helps encourage them to remain all-Apple in the future. Free MobileMe would be an investment in hardware customer retention, and it doesn’t even have to be completely free."
Personally, I'd go one step further, integrating the bottom tier of MobileMe with the iTunes Store, customers of which already have unique Apple IDs. The online storage associated with MobileMe could then be used to secure purchases, say at a 1:10 ratio (1 GB allotment usage for 10 GB of purchases or rentals, which wouldn't actually have to be stored redundantly in your account), in case your machine encountered some disaster and you had no local backup. Premium versions of MobileMe, offering more storage, could cover proportionally more rentals or purchases.
Customers who currently have MobileMe and iTunes accounts associated with different IDs could have both (all) Apple IDs associated with a single merged account.
Given the fiscal need to tie revenue to products and services provided, Apple could lace the free version of MobileMe with iAds, purchasing them itself if necessary to insure that the expenses involved were offset by revenue.
I'd also go one step further with the premium (family/workgroup/professional) version of MobileMe, providing it with some of the capabilities of Mac OS X Server, like the ability to create a wiki or a shared calendar in the cloud, which could be accessed by others with any type of MobileMe account, also merging in the iWork.com service. This tier should also have premium domain hosting and web authoring capabilities, like a professional version of iWeb.
Put enough value under one roof, at a price your customers perceive to be at least arguably a bargain, and there'll be many more customers than if they have difficulty justifying the purchase. This argument is far more compelling for services with relatively high up-front costs and low incremental costs than it is for hardware products with higher incremental costs.
Apple has maintained the price of MobileMe at a high enough level that they should be able to offer a complete, very sophisticated service without raising prices at all, and the cost of operating a basic subset of that service should by this time be low enough to be coverable through tasteful advertising alone, allowing it to be offered for free, with the competitive benefits that Charles Jade outlines above.