Data visualization is one of those things that execs/stakeholders think will be "quick and easy", but turns out to be monstrously difficult and expensive.
I would kill for 35% wrangling + engineering. Finding, validating, cleaning dirty data... that's like 80% or 85% for me. Maybe the estimates here are based on a pre-existing clean (mostly) data source or warehouse.
Great analysis thanks for sharing. Amazed how productive you seem to be!