Originally Posted by UniChris

It's certainly not an efficient design for bikes, as I suspect it treats bike progress through the intersection as functionally equivalent to pedestrian progress. It certainly does not teat bike traffic as roadway traffic, but then it can't realistically treat a mis-positioned through lane as a through lane.

What's the timing of the pedestrian crossing light relative to the main traffic light? My guess would be that when the pedestrian light signals crossing, there's a red for vehicle, quite possibly an all-walk phase. Technically that works, since anyone making a right-on-red is supposed to stop first.

FWIW, some of the worst designed-in conflict situations I've seen are where there are distinct road-adjacent bike and walking paths; drivers like to screech onto the the bike crossing without regard, and then wait there for a gap in pedestrians.

A realization that's growing more and more in my mind is the reality that isolated bike infrastructure gives a sense of safety in between intersections, but only at the high price of a far worse situation (in terms of both safety and convenience) at the intersection.

A problem for planners though is the the cyclist population includes both those with traffic skills and those (both adults and children) without. Sometimes it would be best to expect distinct behavior - interestingly the stretch of Livermore, CA in the original topic seems in recent imagery to have had both an on-street lane and the parallel MUP in much of the leadup to the crash location. But the crash location itself has only decade-old imagery, so it's hard to know if the on-street lane continues.

It's further interesting to note that the leadup to that intersection on Stanely shows the full range of options: a Dutch-style displaced crossing, a bike lane that formally swaps positions with the right turn lane, and then the dump-out of the MUP into the pedestrian-equivalent situation of the incident intersection.