Any moderate speed planing-hull boat will have its speed affected by a change in the power-to-weight ratio. Naval architect George Crouch proposed, and many examples of Boston Whaler boat hulls confirm, that the effect of the power-to-weight ratio on speed is at the 0.5-exponent. Expressed mathematically
NewSpeed = OldSpeed × ( (NewPower/NewWeight)/(OldPower/OldWeight) )0.5
To simply, we assume the weight of the boat does not change. The relationship then simplifies to:
NewSpeed = OldSpeed × (NewPower/OldPower)0.5
If the speed with 500-HP were, say, 40-MPH, we can estimate the speed with 350-HP by Crouch's method to be
NewSpeed = 40-MPH × (350-HP/500-HP)0.5
NewSpeed = 33.5-MPH
Repeat the above as desired with different new and old horsepower to find other predicted speeds when no weight change occurs based on whatever speed you wish to assume for the old configuration.
Experience applying Crouch's method to speed prediction to Boston Whaler hulls has shown very good correlation between predicted speeds and actual outcomes. So there is no basis to alter Crouch's method to calculate the effect for a 2007 270 Outrage, unless the top speed with 500-HP were way off the chart, say 65 to 70-MPH. The effect of the hull characteristics may change somewhat at extremely high planing speeds as less and less of the hull remains in contact with the water, and the submerged propeller and associated gear begin to have the dominant influence.
Without knowing the exact weight of the boat before and after the change in power, the effect of weight change is not applied in the above example. Of course, as weight decreases, performance improves, but this change also occurs at the rate of the ratio to the 0.5 exponent. If you wish to assume there was a weight reduction by a factor of 0.95 (five percent reduction), then repeat the above calculation with the 350-HP figure increased to 350/0.95 or 368-HP. The predicted speed change from the assumed 40-MPH speed would then be to 34.3-MPH.
An additional factor when changing to a single engine from twin engines will be some reduction in underwater hull drag from elimination of one engine gear case and propeller. This will also generally tend to improve performance, but predicting this will be nearly impossible. The overall efficiency for using two propellers to propel the boat might actually be higher than using one propeller is some situations, particularly with outboard engines were the propeller diameter may be limited by the aperture of the engine gear case.
Crouch's method has no factor that accounts for the number of strokes-per-power-cycle for the propulstoin engine, so there is no way to account for the new engine and the old engines being of some different design in regard to number of strokes-per-power-cycle.
Engines of different designs will have different power curves, different engine speed ranges, different gear reductions, and different propeller shaft speeds, and these many factors will require the propeller design to be altered to suit the engine in its application on the boat. The Crouch method assumes the engine, gear reduction, and propeller design and selection are of no influence, that is, that in each case those factors are equivalent for either power choice.