This story has nothing to do with what you mention. Apple can already build what you mention using Google's map data.
This is simply about replacing the provider of maps and the locations databases behind it. This has nothing to do with Maps.app, which is a barely functionning mess of basic functionality on iOS compared to what the competition has to offer.
The efforts Apple is putting towards this back-end infrastructure would be best spent on making the front-end application better. If only Apple didn't operate in a "small business" mode where they have small teams that do many different projects depending on priorities rather than more teams working on more stuff at the same time.
There is greater cohesiveness in small organizations. By having so few employees, Apple is able to maintain the "start-up" culture. Contrast this with Microsoft which employs almost twice as many people as Apple, and Microsoft doesn't have a retail division like Apple (I'm betting the bulk of Apple's employees are Retail). There have been reports that Microsoft suffers from political infighting which have played a role in delaying Microsoft's entry to new markets.
It's also not a very good comparison in the fact that Safari wasn't built from the ground-up by Apple. They basically took years of work done by the KDE Community, forked it, promised to give it back (according to the terms of the GPL), went underground, showed back up a few months later with an uintegratable mess, basically screwing the original developers out of the improvements they made.
Go back to your first paragraph. The Maps.app is a NATIVE maps application, which admittedly doesn't offer the same functionality that competitors' native maps applications do offer (e.g. Android). But there are plenty of apps that can fill the gap. With Safari, you're saying that Apple screwed the original developers over. If Apple implemented features like turn-by-turn in the Maps.app, that would supplant developers who have those kinds of apps in the App Store and put them out of business.