Remember the Dynojet is an inertia dyno so it is essentially measuring how fast a set weight is spun. Faster speed (like a higher gear) will change the result and skew figures. Fortunately @Terry@BMS dyno'd the car in both 4th and 5th gear to show the difference the ER Racing catless downpipes provide.
The 4th gear baseline shows 598 wheel horsepower and 589 lb-ft of torque at the wheels. This goes up to 623 horsepower and 626 lb-ft of torque for a gain of 25 horsepower and 37 lb-ft of torque at the wheels. In fifth gear the baseline is 628 horsepower 637 lb-ft of torque at the wheels. Essentially, 30 wheel horsepower higher. Be wary of tuners who use higher speeds to try to skew figures or those show will show you a fourth gear baseline and then a fifth gear tuned run. It's an old trick that basically every tuner knows. Regardless, the post ER Racing catless downpipes run shows a gain of 9 wheel horsepower and 15 lb-ft of torque at the wheels.
So which is closer to the truth? Well this goes to show how much dyno numbers can vary. The baseline runs were done in cooler weather so that definitely can have an effect on the results. Heat soak should be worse and set in more quickly in warmer weather so that may explain why the 5th gear runs show a smaller difference in the peak figures although the real gains are in the mid-range torque. The main thing to focus on is there are large torque gains in both runs showing the factory catted pipes are a point of restriction and that a good set of catless downpipes is a solid power modification on the F10 M5 S63TU V8.