You can’t run a house on just solar panels — that little detail of using power at night tends to get in the way. Your house still needs a grid connection, and will likely still face a power bill every month.

You will need a grid hookup for load balancing, because solar panels are baseload (for an accessible introduction to the basic terminology of electricity generation, see this piece).

But there exist house designs which are net producers of usable energy for the grid, even after load balancing costs and other overhead. They don’t get a power bill every month. They get a power payment every month. Of course this does not usually cover the electricity cost of an electric car, which might put you back in the red on electricity but save you the gasoline bill.

More fundamentally, I would question the implicit idea of fundamental discontinuities in technological development. The very early Internet really didn’t do anything that a fax machine and a phone would not do if you coaxed them a bit. It just did it faster, cheaper and better. The washing machine does not fundamentally alter the process of doing laundry – it just does it faster and cheaper and (occasionally) better.

Technological development, in my view, is a fundamentally iterative process: You create tools that are a little more sophisticated, to do the same job you did yesterday a little faster, or a little cheaper or a little better. Then you find new and interesting uses for your new tool. And once you are convinced that the interesting new use is more than an idle curiosity, you build more refined, more dedicated tools to do what you are now “already doing.” Just a little bit faster, a little bit cheaper or a little bit better.

Technological progress is firmly rooted in the mundane drudgery of present technology. An industrial culture forgets this fact at the peril of shortly ceasing to be an industrial culture.

Of course they do not build to meet the challenges of the future. They build to meet the challenges of today, and that is a feature, not a bug. We do not live in a word in which the future is perfectly known to us. So we do not know what challenges the future will hold until they arrive (or at least are sufficiently imminent that they may without great violence to the English language be considered the challenges of today).

We can, on the other hand, be reasonably certain that most of the challenges of today will also be challenges in the future. We will still want our homes to maintain a comfortable temperature. We will still want potable water piped directly into our home, and wastewater removed in a hygienic and expeditions fashion. We will still want electricity and Internet access. We will still want places to eat and sleep and play and love.

And while one or two of these predictions may be wrong, you do not know which one(s). So you build to meet the demands of the present, because nineteen times out of twenty a demand in the present will continue to be a demand in the future. And you do not build for the demands of the future, because nine hundred and ninety nine times out of a thousand, your guess about future demands will be wrong.

tl;dr: Predicting new demands is hard. Predicting continued demands is easy.

Second, the rooftop gardens can serve two other purposes not mentioned in the post.
a) Temperature moderation. An earth cover on the roof will keep you warm when it’s cold by making it more difficult for heat to escape and cool when it’s hot by containing (rain)water that will evaporate. And a plant cover on the earth cover helps to keep the dirt where it is supposed to be rather than come off with the first gust of wind. And helps with the temperature moderation too through the plants’ own water management.
b) Related to (a), they will moderate the rainwater load on your greywater system. This is particularly important if you don’t bother to design separate black- and greywater systems and just run everything through the same sewers.

Whether either of these mechanical benefits is worth the hassle of maintaining a topside patch of soil, or the extra up-front cost of fitting the building for it is, of course, another question. One which has traditionally been answered in the negative. But this depends on the particular features of local geography and technical expertise available, so it is not impossible that they might be answered in the affirmative here.

Regarding the four reasons mentioned in the original post, I think we can pretty much rule out recreational purposes based on the layout of the roofs – not a high enough grass to flowerbed ratio, and too many industrial-looking solar panels.

On the other hand, I would not rule out a truck garden at all, at least not for the reasons stated. Growing your own food in your back yard was perfectly economic before the rise of modern mega-farms and their associated subsidies, and there exist not unreasonable scenarios for future economic trajectories under which it becomes a reasonable business model again (any scenario involving rising energy costs relative to median income will increase the subsidies required for mega-farms to be competitive with truck gardens). And no, you can’t just retrofit existing buildings to have gardens on their roof – at least not easily or cheaply. Most roofs – even most flat roofs – are not designed for that, so it would require pretty much taking the roof and quite a few of its supports off the building and starting over from scratch.

I see three potential problems for this project not mentioned in the post. In descending order of importance:
a) They will want to take their 3d models and hand them over to some professional windmill manufacturer or site developer, to be run through their proprietary computer simulations to check the aerodynamics of all those long, straight streets. Because if you build a city full of wind tunnels where people expect to have streets, your project will die an ignominious death.
b) Their building models do not seem to contain back yards. It is my experience that people like having back yards in their apartment buildings. They are somewhat more private than the public park, yet allow you to associate with your neighbours in a way that walled gardens do not (leaving aside the fact that walled gardens in a proper medium-density development like this one are impractical). Not, perhaps, a dealbreaker, but rather odd.
c) They allow regular use of cars on flagstone streets. Bad idea from a maintenance perspective. Flagstone car roads do not age gracefully.