I wonder how the counters treat cyclists cutting across the middle between SW and NE corner. Those counts seem awfully low, I see active users more often than not.
This shows most of the significant inputs and outputs down like 30-50%. The EB-NB turn movement dropped from 919 to 210 in the AM! (and 434 to 167 in the PM). The other three PM left turn movements increased slightly, but negligible changes in the AM.
The only real constant is EB Bow Tr in the AM. So you can make a bit of an argument for the extra lane there, but it's still the epitome of one more lane, bro! Bow Tr streetscape could use work one way or the other, and the new Edworthy access would be good, especially in conjunction with Spruce Dr improvements.
Then if you want to improve the main intersection my radical idea is to kill the left turns off of Sarcee (saving 15-20% of the main light cycle time) and replace them with level ped crossing & U-turn about 600m either side of the main lights. Eliminating those turn lanes let's you reconfigure the Sarcee through-lanes to 3 instead of 2, and the yieldmerge onto SB Sarcee can just become it's own lane. Unlike the Bow Tr U-turn, I'd make these dual turns from the right side where you reduce the angle a bit:
It means both directions of Sarcee have to stop, but if you really rub your traffic optimization neurons together, you can time it to make sense with the main lights (just like they did for Bow Tr U-turn). You'd also need to slow this stretch of Sarcee down to 60, but that's a feature, not a bug.
Perhaps that is just what projects cost now and I am no engineer, but just seems like with a bit more tolerance for lower service levels and a simpler design philosophy we could have achieved 95% of the benefits for much less than $335M!
I think the above would be like 80% of the benefit for maybe 25% of the cost