Extreme saving may well be good for the societies in which the resulting investment happens. But a key idea in effective altruism is that there's far more benefit-per-buck available in other places, where everyone is much poorer. If that's true then the "benefiting society" effect of saving more (or the reverse, if it turns out that actually reduced consumption outweighs it -- I'm no economist and don't know how plausible that is) is probably small in comparison with the tradeoff between giving earlier and giving more to help people whose situation is much worse.