Thanks, the visualization of effects is definitely something we went back and forth on for a while. Ultimately settled on the pie chart just so we didn't spin for too long, and it's five minute implementation with google charts.
Comparing different strains is next on the list for features, have a couple good ideas floating around to make it pretty simple.
If you stick with normalized data (add up to 100%), though a pie chart may be visually appealing, I'd still go with a sorted bar graph. I always look for the most prevalent effect or two, and it takes me a couple of moments to maneuver around the pie chart to find them. A bar graph would fix it. Perhaps even statically ordered by labels so you can easily compare two strains visually.
Also with that in mind, you could get rid of the Top 5 text list, since the data would be similarly efficiently portrayed visually.
Another reason the pie charts probably aren't the best representation is because it implies that the different effects of a strain all add up to a whole. E.g. if the "lazy" effect gets larger, then the "sleepy" effect must get smaller, which we know isn't necessarily true.
Comparing different strains is next on the list for features, have a couple good ideas floating around to make it pretty simple.