Yes, understanding algos is valuable, and implementing them of the top of one's head is a nice party trick, but in the end we're paid to solve problems and it's much faster and easier to identify existing solutions and adapt them rather than reinvent the wheel.

Knowing when and what to optimize is vital.