In residency no limits on HFNC, but if you were much above 12 LPM, then generally were moving to the intermediate care unit (better nursing ratios) if available and getting the PICU fellow to take a peek at the patient (formal vs informal consult was worked out between the senior resident, fellow and if need be the PICU attending). There was no O2 limitation either.
In fellowship, at one of those top 10 children's hospitals that I don't think train residents super well, anything above 4 lpm resulted in transfer to the PICU and asinine conversations from the senior residents and hospitalists along the lines of "well we tried 3lpm, then went to 3 and a quarter with no improvement, then went to 3 and 3/4 about an hour later and they just weren't getting better...I think they're going to tire out and need to be intubated as soon as they get to the unit", plenty of times where 6LPM was all that was needed. Additionally, anything requiring >50% O2 was a PICU call, and sometimes the floor would not accept a kid on 45% because that was "close" to 50%
Now as an attending, in a third location, the hospitalists have weight based criteria, <5kg will take up to 6lpm, 5-10kg up to 10, and over 10kg will take up to 12 lpm. Same O2 limitations and BS about getting close to needing 50%.
All told, nursing care matters more than anything else. And while the data hasn't been the panacea I would love it to be, I'd argue there's a lot of variation in practice that I think skews matters. Working in a location where there are limited resident presence, but very high volume and acuity, I guarantee the value/LOS/and intubation rates would be different than what is published (not necessarily for the better in some aspects, I will readily admit). Every year I easily have at least a dozen patients that I crank on high flow that I would have intubated while a fellow because my attendings preferred that. The flipside is that without the cadre of providers, the weaning of flow on some of these kids gets drawn out. I also think that the equipoise in the data can be read into in multiple ways. If you're a pro HF person, then keeping kids out of the PICU is probably desirable and this is a tool that can do that (depending on the implementation), knowing that if you're actually sick, then your course isn't going to change much. If you you think HF needs to really prove something to be used more broadly, then perhaps a less is more strategy continues to be the mantra for bronchiolitis (just like the rest of it). Maybe that means we don't really need to worry about high RR if you're maintaining an adequate gas exchange, maybe bulb suctioning really is all that is needed. I often wonder what bronchiolitis season was like in the 60's before all of our modern day interventions and what the outcomes were and if we haven't moved the needle, then maybe short of treating actual hypoxia we shouldn't do anything...