Here's the most recent study I could find showing that IOs gain faster successful access compared to IVs in cardiac arrest:
http://www.ncbi.nlm.nih.gov/pubmed/21856044. If someone has a garden hose for a AC that's in cardiac arrest, I say go for it. But if that's not the case, then get the IO to have access and look peripherally when/if you have time. Quick, easy, less complications, and frees up time if your solo medic. Granted the medications we push in ACLS are in question of actually doing anything helpful, but a quick IO and med administrations would be better than a multiple attempt IV with time lost. Got a kickass EJ doc but we only gave 2 epi's in 30 mins.
The IO and tourniquet were comparisons about the "temporary measure" you quoted earlier. I don't believe every patient needs a IO, only in the most dire of situations, I agree with you. But your pt that needs some sort of intervention, may or may not have EJ access available, whereas you have potentially 6 IO sites you could access. Unless they're some multiple amputee pt, then it's safe to say their bones will always be there.
Have to agree with Remi, low cardiac output could def flatten/make EJs disappear. Would rather know I can hit an IO quick to help intervene for sure than hope I potentially hit a spot where a vein use to be, potentially wasting time.