I think total elimination from use is not necessary. But for most events it probably is appropriate for the effect to be relatively small, if not marginal. On the other hand, events by definition simulate things that did historically have enough effect that they've been remembered. So they have to have some effect.
I would be even more supportive of the General's concerns had I not experienced a random event in which seemingly every type of factory and RGO lost 5% of productive capacity. If the argument then is that production increases and decreases shouldn't be used, or should be reserved only for inventions, I don't agree, because Paradox didn't impose such a limit on its own construction of events. (The event I experienced was based on a retraction in the farm economy, and was in no way related to technology or invention.)
On the other hand, if the argument is that these commands can be unfairly powerful in their effects, then I agree, and the question becomes what level of effect is appropriate? (Using the Paradox farm retraction event as a guide, a 5% production capacty loss was deemed appropriate by Paradox to simulate a significant decline. Which is an indication that most events using such a command should make any change something akin 1%, since 5% was used to simulate something fairly profound.) The problem thus becomes one of degree.
(On a related topic, I think that prestige gain can be at least as much of a problem as the General's concerns. Prestige determines such things as access to the world market, and can therefore be more imbalancing than added productivity. Yet I've seen no consensus on what something should be worth in terms of prestige. I did experience a Paradox event in which I hired a composer and gained 5 prestige for sponsering an immortal piece of music. So if a symphony is worth 5 prestige, what should other things be worth? If a political assassination, which no one has any control over or input into, costs what it does, what is a declaration of independence worth?)