If the worst case is that an EV powered by brown coal-generated electricity produces about the same amount of CO2e as a regular oil-powered car, I consider that a win.

I don't. I consider that a loss, because "either petrol or electric cars" aren't the only options. The other option is not to drive your car for that trip, or at all. To get (say) 20% less emissions, you can buy an expensive new car with 20% less emissions, or you can drive 20% less and save money.

Our trip ratio across Australia is something like 85% cars, 10% public transport, and 5% walking/cycling.

As noted by the World Health Organisation [1Mb pdf],

More than 30% of trips made in cars in Europe cover distances of less than 3 km and 50% less than 5 km. These distances can be covered within 15–20 minutes by bicycle or within 30–50 minutes by brisk walking.

The short journeys use up a disproportionate amount of fuel, so even if your six a week 5km trips to the shops are just 6x 5km = 30km of your weekly 200km of driving, they'll make up more than 30km/200km = 15% of your fuel use. This is because engines reach their peak efficiency after fifteen minutes or more of driving, and short journeys involve more stopping - your engine burning fuel for you to stay still is as inefficient as you can get.

as well as short trips, at least two-thirds of trips are discretionary. You can do without them. The following is the data of purpose of journey by car from 1992, the most recent year available for such data for Australia as a whole. [source, ABS]

• Shopping, 25.7% of all trips, 13 minutes average trip time
• Work, 22%, 31'
• Social activities, 18.7%, 20'
• Voluntary & community activities, 9.3%, 18'
• Active leisure, 7.4%, 32'
• Child care, 9%, 13'
• Domestic activities, 5.4%, 16'
• Education, 2%, 22'
• Personal care, 0.5%, 16'
• Passive leisure, 0.1%, 22'

We have here figures for the percentage of all trips taken for that purpose. The average time spent driving each day is 1hr27'. The average time per trip doesn't add up to this 87' because not every trip is done every day; but when the trip is taken, that's the average time of it.

Only about a third of trips (work, child care, and possibly education) are non-discretionary and more or less unavoidable, assuming zero public transport and not able to bike, walk, etc. The rest can be set aside ("passive leisure", driving just for fun) or rearranged for efficiency - shopping from distant shops can be done weekly all in one go, etc.

And of course, many of those non-discretionary trips are for short distances - my woman, for example, is living with me just 5km from work, and cycles on average twice a week (not to save fuel, but for fitness). There's also a bus she could take which drops her off right outside work. And some of the non-discretionary trips are unnecessary, for example this family has four members who all live in the same house and go to the same workplace, and drive in their four separate cars, at least two of which are fuel hogging cars. They start at different times for work, but some combination of "I'll go early and go to the gym until work starts" and "I'll go to the shops after work and wait for you to finish" ought to be manageable to at least make it 2 trips instead of 4.

So of the the 85 trips taken by car out of the 100 trips taken, about 57 are discretionary, and 28 non. Those discretionary ones aren't going to be eliminated, but it seems fair to assume they could be knocked off by a third at least, down to around 38. The non-discretionary trips, a combination of cycling and public transport and just generally having your shit together ought to knock them down a bit, let's be stingy and call it a quarter, down to 19. So now car trips are 38+19 = 57 of the total, down from 85. This would put the walking/cycling trips up a bit from 5 to 15 or so.

In this way, we get around a one-third reduction in fuel burned by cars, and improve the fitness of our rather obese nation.

Here in Melbourne, even the train network operator Connex admits they could double the number of trains running without any spending on infrastructure (just on the trains themselves, and the drivers).

While the trains here are overloaded, buses are frequently empty. That's because they run infrequently and are often late due to traffic. As purely private operators, they suffer no fines from government or public scrutiny of their performance, which as a result is crap. And no-one is responsible for co-ordinating timetables, so that for example when you get off a train a bus is waiting, or vice versa. A few painted-in bus lanes and two or three competent managers and timetablers appointed could easily handle this.

So it doesn't seem to be too ambitious to say we could double public transport's share of all trips from 10 to 20. Remembering also that the walking/cycling trips would have gone up, we end up with a ratio of

57 car trips : 20 PT trips : 15 walking/cycling trips

or 62 car: 22 PT : 16 walk/cycle

which is a considerable improvement on the current 85:10:5. And all it took was some effort and thought from the public, and a bit of spending on extra trains and a few competent managers. Which could all be achieved in a few years at most, compared to changing the entire car fleet to electric which would take... how long? And cost us how much? And require how much more spending on roads?

And as Matt says, we in Australia need to knock over our stupid coal-fired stations. That in our "sunburnt country" we don't have solar PV and thermal makes as much sense as Barbados refusing to have tourists. Once we burn the coal it's gone forever, but there is no prospect of Australia becoming less sunny in the future. You have two bank accounts, one has $100 million but no interest, the other has a $10 billion principal which you can't touch because it's 93 million miles away, but it offers $10 million a month interest, which bank account do you draw on?

Cars kill 1.2 million people worldwide annually. Wars only kill 200,000. Even if cars ran on pretty girl's smiles and their exhaust gave us vitamin C, they'd be a bad idea. As Lester Brown notes,

Each U.S. car, for example, requires on average 0.07 hectares (0.18 acres) of paved land for roads and parking space. Thus for every five cars added to the U.S. fleet, an area the size of a football field is covered with asphalt. More often than not, it is cropland that is paved simply because the flat, well-drained soils that are well suited for farming are also ideal for building roads.

[...] the U.S. area devoted to roads and parking lots covers an estimated 16 million hectares (39 million acres), almost as much as the 20 million hectares that U.S. farmers plant in wheat.

Whatever they're powered with, cars consume vast amounts of resources and land. As I said, it's a tonne of metal and plastic to transport 1.5 people. This is abombinably stupid and wasteful.

And they make our cities fugly. Below are two city scenes, one designed around cars, one designed around people on foot. Which would you rather live in?

The places with the highest real estate values for residents, and with the most tourists, are those with few or no cars, good public transport and walkable neighbourhoods. It's what people want. It's just that city planners and home builders are too lazy to give it to them.