You can go round the house turning things off/on to see what's using the most energy with a standard meter.
If you do that with either a smart meter or a dumb meter the most likely outcome is a load of meaningless misleading rubbish. Most of the big items like fridges, freezers, cookers, washing machines, etc have thermostats and other switches that vary the power. If you measure the instantaneous power at any point it's telling you nothing about the mean power averaged over a day, or complete use cycle.
If you wanted a meaningful measurement of an appliance's average power using the electricity meter, you would need to switch all the rest of the house off for a few hours whilst you do it, to prevent the other appliances confounding the measurement, which is anything but quick and convenient. To measure one individual appliance at a time, by far the quickest and easiest way is with a plug-in power meter.
The display for the smart meter can be carried around the house, and shows instant energy use.
Instantanous measurements aren't of any help, see above.
We are with Octopus, on their agile tariff
Once we have tariffs varying in real time according to load on the grid, it'll be the customers who'll need to be agile if they can't plan their day, and meal times, without first consulting a fuel price forecast.
Until you realise that my house was on a small estate - so all the houses were built at about the same time - in fact mine was completed after most of the neighbours (long story!!!!) so teh meter was newer than theirs.
And yet it was only mine that needed urgent replacement as it was too old
They're only replaced on a sample basis, not all of them.
switching them off and on again causes damage
You might as well do that with every other electrical appliance as well then.
when first switched on they don't sound as good
You should have a word with Prof. Richard Wiseman, he's a psychologist who specialises in blind trials debunking 'connoisseurs' like art/wine/HiFi buffs who claim they can tell the difference between one product and another.
100W will only use 1/10 of a unit per hour
Or to put it another way, a 40% increase in my electricity consumption.
if suddenly they start to export huge numbers it's easy to prove, showing past data.
Easy to prove is not the same as easy to sort out. When I had a billing problem it was easy to prove, but it dragged on for about a year, and didn't end until after I'd had a letter from a debt collecting agency threatening to take me to court.
they believe it will help in their aim of reducing overall energy use
The government were forecasting an energy saving of 2.8%, but Dutch researchers found that the actual savings were about a quarter of that.