Each table represents a hypothetical scenario, which is made up of multiple reviews, each made up of a score and a review time. I recommend reading the article on "How to write your own spaced repetition algorithm" to learn more about what each value represents.
Basically, scan the Scores column from top to bottom and see the effect on the Next Due column over time.
Review Time is assumed to be on the next due date, although there are some scenarios where I've simulated a case where the review was either late or early.
Scores range from 0.0 to 5.0, and follow SM-2's scoring standard (also known as the 'q' or 'quality' of the response).
Everything to the right of the arrow is the "output" of the spaced repetition function: n, e factor, and interval. Each time a card is reviewed, the card's ease factor (e factor) is updated. If a card was easily recalled, i.e. higher in the 5-point scale, its ease factor is adjusted higher. Likewise, if it was difficult, the ease factor goes down.
The "Next Due" column indicates how much time must elapse from this review until the next one. Note that the "Next Due" is actually internally represented by an "interval" floating point value, which represents how many days from the current review time the next review should happen.
Finally, the n value indicates how many times this card was remembered correctly in a row.
Everything to the left of an arrow represents the "input" into the spaced repetition algorithm, including its current state (its n, ease factor, and interval).