Although a few sentences have already been written about arms, describing the legs action from the biokinetical point of view appears to be more logical as a consequence of the previous chapter about the main body. It is also much simpler to do since the legs are the base of the whole human body and the only link between us and the ground.
The evolution of species proves that the gravity is not such a huge obstacle to force to engage all distal parts with an aim to keep the main body safe, stable and balanced. A pair of "well-designed" legs is enough to enable us to make a huge variety of motions. If we look what is the main difference between us and animals (except Primates) we can see that our feet can be entirely used for a contact with the ground. This fact have pros and cons - the main advantage is that we do not usually need our arms to help us to move or maintain balance, the main minus is that we are not very flexible in a 3-D movement comparing to almost all mammals.
The impact of this fact is crucial when talking about a golf swing motion where the most important goal of the legs is to maintain balance (and not e.g. to jump or change the orientation of motion fastly). Our feet are a very solid base for the rest of the body, and, what is equally important, they are very sensitive for changes in the motion. They can be like this since the anatomy of a human foot enables to react immediately to the brain's signals with a very rich variety of actions (dorsiflexion, plantarflexion, abduction, eversion, inversion, supination, pronation, etc.) because the lower joint (malleolus=ankle) let the feet move practically in all possible directions. The good thing is that there sould be no special attention paid to the ankle joint and its limitation since the gravity and necessity of maintaining balance does it for us. It is more important to look at what feet are supposed to do (or, better said, how they should be placed) and what impact it may have on segments that are located further from the ground.
The other legs joint of huge importance is the knee joint which is, BTW, the biggest joint in the human body and is homologeus to the elbow joint in the arms, however, they both are acting differently as regards planes of motion. Simplyfying the topic, the knee joint can act in two independent planes - it can flex/extend as well as rotate. What is interesting, a two-axial joint always have the primary and secondary motion, i.e. one of the motions are much frequently used from biomechanical point of view.
In case of the knee joint, flexing and extending are the motion of the primary importance - but only in one direction - knees cannot be bent backwards. That is why the ankle joint is not located in the middle of the feet.
The lateral rotation is very limited (up to 30 degrees only) - and this fact is crucial for our golf swing motion since the knee rotation happens exactly in the same horizontal plane. Another fact worth mentioning is that flexing/extending action of the knee is strictly related to the foot motion, that means it can be subdued to a proper feetwork. I want to underline the word "feetwork", not "footwork" - the greatest ballstrikers never let put the work on only one foot practically until approaching impact (well, the truth is that if the stick is long enough and the stance very wide the rear foot is too far away from the CoG at impact and it is not entirely flat on the ground at impact).
Lastly, the hip joints, that are the true links between the main body and the legs. The fact that there are two of them plays a crucial role in all CoG shifts of the human body in action, as well as determines which of the two hip joints is the protagonist. When a main body equipped with two lower distal parts is turning and wants to do it effectively, it must act as if we have only one leg. Independent two joints are the obstacle in a turning motion and a bliss for a sway. This is why the primary thought of a golf beginner is to sway with the hips in order to gain power and not to put the vast majority of the weight of dynamically turning mass on one leg. Why ? because our brains see the ball flight as a linear motion (from the point where we are to the target). When the same beginner begins to believe in gaining power from rotation of the body, he instinctivelly places the CoG on one of the hip joints. Moreover, it is much easier to understand why it should be the lead hip joint, and not the rear one - the orientation of the linear movement is just this way. Combine the linear movement concept with conviction that rotation gives the true power and the effect will be a great CoG shift in a hip area. It is that easy, IMHO, and in a perfect accordance with the optimal hip area movement described in the previous section.
OK, now ad rem. The golf swing motion relies generally on turning the whole body back and forth. Thus, we need to find limitations in joints to ensure automatism in the whole motion. Since we already know that the main body should encounter the sum of small X-Factors it becomes obvious that it needs a stable base. Neither feet, nor knees should enable the rear hip slide back and lose the coil.
Therefore, the most important limitation that is to be found is to make impossible for the knee and ankle joint to move separately from the rest of the upper body. Unfortunately, the dominant motion of the knee joint happens in a perpendicular plane in relation to the orientation of the golf swing motion, and nothing can be done about it. That's why there are a lot of golfers (even the greatest ones) who squat down or up in certain moments of swing. It's inevitable in such a dynamic motion. Luckily, our brains are able to control it when focusing on the primary goal which is hitting the ball and not the ground. That is why casters (early release golfers) stand up before impact and that is why great golfers squat after transition in order to maximize the optimal velocity of the clubhead. Please also note that the small feet CoG shift reflects in the knee joint movement - the more is the weight set to the heel, the more the knee extends in a rotary motion, and viceversa.
Nevertheless, a lot can be done in another plane, i.e. in finding the limitation of the rotation in both knee as well as in ankle joints. Proper feet stance is crucial because there is a significant amount of friction between the sole and the ground. The more the feet is directed towards the target of the motion the more limit will be found in both joints. That's why Hogan moved his heel of the rear foot back, setting it at least perpendicularily to the target line - which move is well documented in a lot of films as well as in his books. It makes finding the limits of knee and ankle rotation very easy. When there is limit, there is no room left for screwing the motion. IMHO, it's enough, and although I follow the Sevam's thread carefully as a big Hogan fan, introducing another thought as e.g. screwing the foot into the ground is not necessary since it puts the focus on timing. The rear foot is "screwing" into the ground automatically when the knee rotation finds its limits because the foot cannot turn back due to the resistance. Therefore, in my opinion, presetting the torque does its job perfectly without necessity of thinking about it. Funny thing is that the knee joint rotates only when it is bent what would suggest that one should hit balls with straight legs...but such a swing would be deprived of power, thus, it is not a subject of the studies.
The left foot, as we know from "The 5 Lessons" should be flared and this concept is also biomechanically very sound. It allows to find the limit of rotation of the lead knee even faster than the rear one. The knee that cannot rotate further bends inwards what almost automatically makes the weight shift on a lead hip joint when the linear CoG shift happens. If the lead foot is put inward, instead flared, all tension in the knee and hip is lost. Moreover, the sum of small X-factors is being achieved properly when the lead hip "stays" early in the backswing allowing the rear shoulder find its limit without making the backswing too long and loose.
Lastly, it is to be pointed out that setting the weight in various patterns influences the limits. If a swing pattern requires setting more weight on a front foot, the rear foot needs to find its limit later in the motion because of the upper body position (mainly a necessary spine angle lateral bent when swinging back); thus, in such a situation, keeping the rear foot perpendicular or even slightly turned inwards is not a great thing. The same is true when, hypothetically, one needs to keep the majority of the weight on the rear part, but, as explained earlier, should not be discussed there because of the linear orientation of the swing motion.