Public transport in the age of the smartphone

Summary

Now that most people have smartphones, why do we need fixed timetables for public transport?

Detail

My wife and I visited Sydney a while ago. We haven't spent much time there and don't know the way around well. To navigate, we used Google Maps and selected public transport as our method. It is quite amazing -- Google is able to tell us where to walk, which bus/train/tram stop to wait at and it  even designs connections allowing time to walk between them. In my experience (and I now use it exclusively when planning a journey via public transport) its performance has been flawless.

It got me thinking -- 85% of Australian adults now own a Smart Phone. If we reach a situation where people can be relied on to have a smart phone, how can we take advantage of this to (re)design public transport?

Imagine

It's 2017, and you want to go somewhere by public transit. You use your smartphone to plan where and when you want to go (whether it is now or in 30 minutes' time). The app lodges the journey request with the local public transport authority (PTA), where there is a computer that dynamically generates the routes and timetables to minimise waiting times and maximise the directness of routes. The transport fleet is different -- there are still many large buses/trains/etc but in addition there are now a local fleet of small mini-buses (say 8 - 12 seats) that facilitate relatively local movements or act as feeders to the larger/faster routes.
The novelty of this idea is that the routes and timetables are dynamically generated -- there are no fixed routes or timetables -- these are designed on-the-fly in response to demand. There are New Years eve fireworks in Sydney and 1.5 million people want to go home at 1am? No problem -- feeder shuttle buses take people to the trains and larger buses take people to places where rail service is poor. Because the route planning is centralised, it can even avoid bus-jams (traffic jams comprised almost entirely of buses) by rerouting accordingly.

From the users' perspective, the service would operate similar to using Google Maps to plan a journey.
From the drivers' perspective, it would be like using a GPS navigator that tells them where to go and whom to collect/deposit.

Why?

Public transport is hugely under-utilised -- many buses/trains are well below capacity. Yet one often needs to wait for service. A system such as this would be much cheaper to roll-out than additional vehicles, and would result in a leaner service that was much more agile and efficient.

How is this different to Uber?

(disclaimer: I've never actually used Uber)
There are a few ways I can see

1. It is designed with systemic efficiency in mind. Because the drivers are employees and can be relied on to work particular times, the central planner organises them collectively to move people. This will achieve much greater performance than a system where individual drivers compete for individual passengers. This point can't be overstated -- as a hybrid system, performance would be excellent.
2. It's publically owned/operated. I would envisage that this is state-subsidised. Also, because journeys are shared (just like a bus/train), trips would be very cheap compared with a taxi service.
3. It still uses connections to maximise efficiency. Of course, because connections are planned, you would never miss one. For example, your shuttle bus deposits you at the train station, the train waits until people have boarded and then departs. There is no timetable.

Problems?

 There are two (related) problems that I can see

1. 15% of Australians lack a smart phone [1]. A service such as this would exclude non-smartphone users. I think a government subsidy to give all low-income people a smartphone would still give savings in public transport alone if this were implemented.
2. Technology barrier. Old people and people with sensory or cognitive impairment could tend to be excluded. This would need to be carefully managed to ensure access.

----------------
[1] This number is changing rapidly. This article states that 30% of Australians who lack a smart phone are planning to buy one in the next 12 months. If that eventuates, by 2017 only about 7% of Australian adults would lack a smartphone.