with the run duration, things you thought will change. by video of your run the hopping your crankshaft is bouncing back and forth, which transfers to piston tops oscilating, not steady rotation like you expect. saw data/video on offshore inboard racing engine(early 80's), crank/cam position sensors, almost 1/4 twist on crank coming out and going in to water, similar amout reflected to cam. overfueling on continuos pull is different than short burst pulls, heat soak and rejection start to switch places and softer metals will give(piston tops), an aluminum head expands and valve seats migrate with valve movement, thermal attraction of like metals and temp ranges. over/under flow of coolants(air; intake and cooling, fuel, oil and water) and you won't get good thermal transfer, look at high performance piston aircraft engines they had louvers that moved during various temp of engine to control heat transfer. when looking for examples when considering component mods, look at high performance diesel marine engines, example; Isotto-Franchini, scarab(ford tractor) marine merlin=300HP Inline6. Paxman out of england makes larger engines, henry ricardo ran this company's engineering from 1930's to 50's. on salt for hp calcs, can't remember what they refered to it as but i think it was frontal horsepower load, similar to planning hull to displacement hull boat, 10% increase requires about 2-4X power, depending on variables. good luck
