|A "perfect" lunge with the knee behind the toes.|
|A lunge with the knee well past the toes.|
The concern is generally phrased that letting the knee go past the toes will damage the knee in the long run. Others will argue instead that as long as the knee is strong enough that it shouldn't matter.
From a technical terminology standpoint we are talking about the amount of flexion at the knee joint. The top picture shows about 110° of knee flexion (0° is the knee straight). The bottom picture would be closer to 130° if the thigh were parallel to the ground - as it is the angle looks similar but this fencer has their hips higher than the other.
The forces on the various structures of the knee change throughout the range of motion. The ligaments are tightest around 90°, so the lunge differences we are discussing won't matter since both options flex the knee past 90°.
Compressive forces on the cartilage and articular surfaces increase with knee flexion. Forces between the underside of the patella and the femur are larger than the forces between the femur and the tibia and menisci. Theoretically this means the kneecap would be at greater risk, but really there are enough differences between the ways an injury can occur that a simple comparison of the numbers is not compelling.
Well, I couldn't find any published research on this question. But since all the peer-reviewed research on fencing could fit on a 3.5" floppy disk, that isn't surprising. The closest research I could find was on squat depth, such as, "Are Deep Squats a Safe and Viable Exercise?" This is a commentary article where a for and against position are presented by different exercise scientists. The article summarizes a large number of relevant studies. I'm going to extrapolate from this to our question about lunges - any time conclusions are extrapolated or generalized from separate research we should be aware that we could easily be wrong. There could be any number of unaccounted for factors in our lunge question that just don't show up in squat research.
There is ambiguity in the research regarding the risk of injury from deep squats. Overall it would seem that there is no clear correlation in population studies between injury and squat depth. But this could be explained by self-selection. That is, those individuals who continue and/or do well with deep squat exercises are those whose knees would not have been injured - for whatever reason, perhaps some natural advantage of their joint architecture. While those who refrain from deep squats, even for unconscious reasons, may be more susceptible.
Biomechanics studies clearly show greater compressive forces from deep squats. And greater forces do mean greater likelihood of both acute and chronic injury. So the concern is certainly plausible.
The above is for healthy individuals. In those with known knee problems we can be confident that limiting squat depth, and by extension the knee position in lunges, will reduce pain. The safe generalization is that the motion - squat or lunge - should remain within a pain free range. Whether that is sudden pain during the motion or the ache afterwards doesn't matter, pain is our indication to avoid that extreme.
There is also, of course, the tactical considerations of lunge length. The more the knee is flexed the longer the lunge that can be obtained. A longer reach is it's own tactical advantage. However, a longer lunge is also a longer recovery. So there is a trade-off - higher risk of getting hit by the riposte or afterblow. As a technical coach I would counsel a more conservative lunge, but that is based on my study of Hope's system.
As a fencer I will sometimes take much longer lunges, when I feel the opportunity is right. Sometimes I am wrong and my opponent gets the afterblow. Sometimes I still fall short, but don't get hit either, in which case I was still wrong, just not as wrong. And sometimes I hit where otherwise I would not have and do so safely.
The ability to recover quickly from a long lunge is dependent on the strength of the lead leg, and I think this is where the biggest component of strength actually plays into our question. A strong leg is less likely to suffer many kinds of injuries, but not really the kinds of injuries that occur from deep knee flexion. There are legitimate exceptions though. One of the fencers I train has healthy knees that can withstand a long lunge, but her strength in balance is lacking and she falls too often. I'd tell her to limit her lunge length to reduce injury and falls (but she's 14 so I don't expect it to take).
On the other hand one of my frequent opponents has a nice technically perfect lunge length and knee angle, but her strength-to-weight ratio isn't as good so she has a slow recovery. Which I can exploit when we fence.
If the historical system you study shows a deep lunge then I would feel free suggesting a person train that, as long as it's pain free. In general, I will always counsel a conservative approach; it is easy for young, healthy athletes, or simply highly competitive adults, to overtax themselves in ways that they will regret years later.
And there is a component of individualization, rather than a one size fits all answer. This is the responsibility of the coach and requires time, observation, experience and thought to answer
1. Schoenfeld, B., & Williams, M. (2012). Are Deep Squats a Safe and Viable Exercise?: Strength and Conditioning Journal, 34(2), 34–36. http://doi.org/10.1519/SSC.0b013e31824695a3