Presumably they mean that there are efficiency losses in charging the supercapacitor banks used to fire the lasers; so that if you consider the system over multiple duty cycles rather than over a single cycle, it's no longer energy-positive. (I.e. if the system were capturing its emitted energy — and that emitted energy needed to be enough to act as a grid power source feeding input power to the supercapacitors, rather than merely being the equivalent of the direct output power of the lasers per shot — then it wouldn't be enough to sustain the reaction.)
But personally, I don't know whether that's actually important. Power plants usually consume a nontrivial fraction of their own produced power to power themselves, and in fact consume more than 100% of produced power when starting from a full stop — meaning that in initial few-shot conditions, even when feeding back their own produced power into themselves, they still need (huge amounts of) external power input to get going, like a car engine needing a battery + starter motor. Only a rare few kinds of power plant can be used to "black start" a power grid. Most types of generator need to overcome initial higher resistances, e.g. inertia (and thereby back-EMF resistance at the transformer) in getting heavy turbines spinning from a stop.
It wouldn't be at all strange if a practical fusion power plant turned out to be energy-negative over a few-shot run (i.e. required "bootstrapping"), but then became energy positive over a theoretical 24/7 run at whatever its optimal duty cycle is. And a single-shot run becoming net-positive would be a good point to start to consider those more practical calculations, since they'd have been useless to consider until then—a power plant can't possibly be net-positive over any kind of runtime + duty cycle, if its core reaction can't be net-energy-positive when considered in isolation.
Which is, to me, why it probably does make sense for ITER to be excited. They've reached the point where they can stop using a lab-bench model of power efficiency, and start trying to come up with another, more full-scale model of power efficiency to replace it with.
But personally, I don't know whether that's actually important. Power plants usually consume a nontrivial fraction of their own produced power to power themselves, and in fact consume more than 100% of produced power when starting from a full stop — meaning that in initial few-shot conditions, even when feeding back their own produced power into themselves, they still need (huge amounts of) external power input to get going, like a car engine needing a battery + starter motor. Only a rare few kinds of power plant can be used to "black start" a power grid. Most types of generator need to overcome initial higher resistances, e.g. inertia (and thereby back-EMF resistance at the transformer) in getting heavy turbines spinning from a stop.
It wouldn't be at all strange if a practical fusion power plant turned out to be energy-negative over a few-shot run (i.e. required "bootstrapping"), but then became energy positive over a theoretical 24/7 run at whatever its optimal duty cycle is. And a single-shot run becoming net-positive would be a good point to start to consider those more practical calculations, since they'd have been useless to consider until then—a power plant can't possibly be net-positive over any kind of runtime + duty cycle, if its core reaction can't be net-energy-positive when considered in isolation.
Which is, to me, why it probably does make sense for ITER to be excited. They've reached the point where they can stop using a lab-bench model of power efficiency, and start trying to come up with another, more full-scale model of power efficiency to replace it with.