While the physics do not change, there are many variables:
1. "Broken legs" - there is a tremendous amount of legroom in my car, versus little to no legroom in a subcompact (Hyundai Accent, et al). If legs are broken upon seat impact, it would take quite an impact to do that in my car.
Do you have enough room for your child(ren) to fully straighten their legs, plus an additional 6-12 inches beyond that? During a crash, there is a tremendous amount of movement, in part due to belt stretch. The only vehicles that could possibly have that much room would be three-row vehicles with the second row removed or stowed. Possible, but not common.
2. Collision-mitigation - to my knowledge a 1999 Toyota Corolla does not have "CMBS". My car does (tied to laser/sonar cruise control).
Collision mitigation brake system doesn't do anything to reduce the likelihood that you will be rear ended, hit from the side, hit road debris, or drive off the side of the road for some reason. It also doesn't help you any when another vehicle enters your lane. CMBS does nothing to help protect the vehicle occupants from injury when the vehicle is involved in a crash, aside from activating the seatbelt pretensioners, which isn't relevant for children in harnessed seats.
Crash avoidance systems are a good thing, but nothing can protect you from every crash. The purpose of seatbelts, airbags, child restraints, vehicle safety cages, etc is to protect the vehicle occupants when a crash happens.
3. Suspension travel and brake capacity (to keep car solid and/or provide buffer as to not "glue to the ground" and cause further injury)
My point is - there are many variables which exist. I didn't want to stir the pot, but taking these studies at face value is like saying "every [insert demographic type] is stupid."
Cars are complex machines, and safety (including safety cages, etc.) has improved drastically over the past 10, 20, and 50 years.
The studies do NOT break out safety on a manufacturer (or model) by model basis.
The best I've seen is at IIHS.org. Compare these three:
1. BMW 5-series:
http://www.iihs.org/ratings/rating.aspx?id=1374
2. Hyundai Accent:
http://www.iihs.org/ratings/ratingsbyseries.aspx?id=586
3. Toyota Corolla (pay attention to 1998-2002):
http://www.iihs.org/ratings/rating.aspx?id=87
Yes, I do believe that newer vehicles are in general safer than older vehicles. When looking at vehicles of a similar age/era, there are some that are safer than others. None of that changes the fact that a forward facing young child is at risk for internal decapitation due to neck shear - the child's skull is literally ripped off the spine. Rear facing supports the head, neck, and back in alignment throughout the crash, virtually eliminating internal decapitation as a risk.
No matter if the child is riding in a so-called "domestic deathbox" or a 2012 5-star rated vehicle, accident physics remain the same. Infants and toddlers have disproportionately large heads and underdeveloped spines. The best protection you can provide for your child is to drive a properly maintained vehicle; secure all occupants properly; secure all cargo properly; drive the safest vehicle you can; follow all the rules of the road, especially leaving a proper stopping distance; be a defensive driver.
You referred to a RF vs FF debate. As far as I know, no such debate exists when looking at the question of what is safer. Nobody who knows anything about passenger safety believes that a child is safer forward facing.