How do latent and observational learning force a reconsideration of learning views offered by classical and operant conditioning theorists?
This is an interesting question, and the difference lies mostly in the means by which a person learns something. In classical and operant conditioning, the learning occurs through an actual, live, active experience that is happening to the person themselves. So, for example, let's take learning how to change a light bulb. Latent and observational learning have a person not actually changing the bulb to learn the skill; rather, they just watch someone do it and thus are able to pick up the skill themselves. With conditioning, that person would have to actually go change the light bulb and experience reinforcements or punishment through the process in order how to do it right.
For example, if the person tries to take the old light bulb out but squeezes too hard and breaks it, cutting themselves, that would fall under classical and operant conditioning. It would be classical conditioning if the next time a person approaches a light bulb, they feel nervous. It is operant conditioning because they learned NOT to squeeze the lightbulb too hard, because they were punished the last time they did; their behavior was shaped by that reinforcement.
With latent and observational learning, there is no hands-on learning experience. You see tons of people changing the bulbs, and are able to do it correctly yourself the first time because you saw it modeled. That isn't the case with conditioning; in conditioning, you experience and learn for yourself through a series of stimuli and reinforcements. Conditionists would argue that all learning comes through experience; but latent and observational learning shows how that is not always the case--we can, indeed, learn things simply through watching others.
I hope that helps; good luck!