KRE D-port heads, 85 cc vs 96 cc on the 7K3 heads. I haven't checked their numbers but the article says 9.2:1 for the iron heads, and 9.8:1 for the aluminum, .6 difference, which would barely cover the heat loss into the aluminum. That makes the thermal efficiency very close for both engines.

The airflow data that was included averaged 10% more on the intake across the board and a little less for the exhaust. At .500" lift, the numbers were 7K3 - 211/162 cfm, KRE - 231/170 cfm. Valve sizes were 2.11/1.77 for the 7K3's, 2.11/1.66 for the KRE heads. At the "important" lift points (60% of max lift for the intake, 90% for the exhaust) the flow was about 10% more for both ports. The torque increase and the hp increase were also about 10%, no surprise considering that the engine is basically an air pump. Torque at 2500 rpm (starting point) was 450 for the iron heads, 465 for the aluminum heads, so no low end penalty for more airflow either.

The improvement isn't all due to the airflow, however. The timing needed for the best performance with the iron heads was 37 degrees, while the KRE heads with the revised chamber shape and better spark plug location only wanted 33 degrees. That points to a quicker burn and less "negative" work done as a result.

One other thing to keep things honest, they ended up using roller rockers (same ratio) on the KRE heads, which could account for up to 10 hp (but usually half that) depending on how bad the stock rockers were.

Cam was decent, with 230/236 @ .050" and .488"/.491" lift (Comp Cams actually), headers were a bit small at 1 5/8", no length mentioned.