Why does the author say that we would need to know the shape of the slide to find the time taken for the child to reach bottom of the slide?
As you've discovered, the speed going down a frictionless slide only depends on the vertical distance. This speed is not the vertical component of velocity. It is the magnitude of the velocity. The vertical component of velocity will be less than this on an inclined slide.
To make the geometry as simple as possible, I'll look at inclined ramps (no bumps, no curves; just a ramp at some angle inclined at with respect to horizontal). To keep the numbers simpler, I'll use g=10 m/s2 rather than 9.80665 m/s2. Suppose the slide has a vertical drop of 5 meters. That means the velocity at the bottom of the slide is 10 m/s. The average velocity is half that, 5 m/s.
Now let's put different length slides in place. A slide that is 5 meters long means you are falling rather than sliding. It takes one second to drop 5 meters. What if we used a ten meter long slide (i.e., inclined at a 30 degree angle with respect to horizontal). The velocity hasn't changed, but the distance has doubled. It takes two seconds to slide down this slide; twice as long as the vertical drop. Use an even longer slide, but still a 5 meter vertical drop, and it takes even longer to get to the bottom. With a 50 meter long slide (5.74 degrees with respect to horizontal), it takes ten seconds, or ten times as long to get to the bottom compared to the vertical drop.
In general, the time needed to reach the bottom of a frictionless inclined ramp is given by $t_\text{slide}=\frac l h t_\text{vert}$, where $l$ is the length of the ramp, $h$ is the vertical drop, and $t_\text{vert}$ is the time it takes to fall that same vertical distance.