Even ignoring the poor arithmetic, they could still have a significant effect if the rollout is managed such that you put them in problematic areas first. I assume there are places where the grid is under greater strain than others, and installing meters in those places first may well give you most of the control you need well before all 100% are fitted. It's quite possible that 30% of meters done, but inthe right places is all that's needed. I don't know, but it seems eminently plausible.
I can see two problems with such a targeted approach - one practical, one political.
The political one is this - what's the likely effect if only certain areas are affected by rationing*, while others aren't ?
* Whether that's "strong financial incentives" to change usage, or actual disconnects to cut demand.
My feeling is that if you start doing that sort of thing, those affected by higher peak rates and/or disconnects are going to be "quite unhappy" when they know that 2/3 of the country aren't sharing their pain.
The practical one is, it's going to need a lot of people to swap all these meters. It makes sense to have them spread all around so they can mostly work local to their home. This means having a spread out rollout - ie one person does an area over a longer period, rather than drafting in 10 to do it quickly but having to put them up in lodgings. I think that once you have to be putting up large numbers of people in accommodation then the costs will go up somewhat.
Added to that, there is the fundamental issue that a lot of the gains won't come until attached loads within the house are "connected". That means having appliances etc able to respond to the information the meter provides. They aren't going to appear on the market in huge numbers until smart meters are ubiquitous, and they'll only get installed as part of a normal replacement - eg I'm not going to go out and spend £300 on a new fridge just to save maybe a fiver** a year off my leccy bill, and I'm hoping the fridge I've just bought will last long past the end of the planned smart meter rollout.
** Figure plucked from air, but if government are only talking of saving £23/yr total per household then I think a smart fridge is probably not going to save much at all.
Of course, government could mandate certain stuff - for instance it would make sense to mandate that all new appliances after a certain date had to be "connectable". Manufacturers might see it as an opportunity to charge a premium on the price (they always seem to, and in any case they'll have additional development and support costs to cover), and buyers will see it as being "shafted again" by "green taxes".
My feeling is that you'll find a small proportion will go out and do it anyway, the majority will sigh at what they consider yet another excuse to have their wallets emptied. I'm afraid that in a large part of the population there is this perception that it (climate change, renewables, etc and all that goes with it) is all a big excuse to add taxes.
