Are these figures factual? I assumed the orphan chance to increase but only by a very small amount, small enough to completely ignore it. Does adding 500KB to a block really increase the orphan chances by 0.4%? I thought it would be many orders of magnitude less.
I think it was Gavin who estimated the orphan cost per kB, and it was not insignificant. I can't find the post off hand, but perhaps someone else can point us to it. I'd like to be reminded of the actual numbers too.
I do know that the core devs want to reduce the orphan cost by propagating blocks by TX hash whenever possible. High orphan costs are sometimes blamed for the hesitation of some miners to build larger blocks.
Anyway, thank you both, I'll need to do some more reading/investigating on this. I'll check back here tomorrow whether someone did find that post
