I doubt it. I don't know how many people went and bought an iPhone 5s because of iOS7 even though it was a huge graphical update. It would take something very special for Android L's software alone to move units en masse. Not to mention that the only people who really get to notice a new Android update are techheads and Nexus owners, who are already converts.
That we know of. We don't know what OEM integration and updates will be like, because they usually have the excuse that the latest version is sprung on them only when it is launched. Now, they have had some time to make inroads into deploying it sooner and across a greater spread of devices, and it will be difficult for them to abandon devices now. KitKat already lowered the hardware benchmark, and it's no coincidence that it is the version of Android that has made the most progress in terms of install base.
I do see this trend (Android to iOS switching) being most prevalent in the US, where the cost of owning an iPhone is comparable to the cost of an equivalent Android device. Outside of the US, I don't think so.
There are a lot of iPhone users that have purchased the device because it works, and a lot that use it because of the fluidity and responsiveness of the OS. Some of those may find Android L to be good enough on those fronts to migrate over to.
I also think that we are perhaps overstating the amount of users on Android that are simply there because of Apple's previous screen sizes. So I think that in the end, it will be a pretty even split, with the spread being different per market.