Breaking up larger Asks in smaller parts, working with clear deliverables and milestone payments has been proven again and again to achieve better results and works especially well in an adversarial, experimental environment.
You definitely have a point with funding the right initiatives but r/r for Optimism differs from venture funding.
If we were not „bound“ to Optimism governance guidelines, we’d also advocate an extremely focused grant funding approach, for example on truly unique projects & top teams with tremendous network effects.