For the lack of quantifiable dynos at the moment (save for the JDL dyno on their website), I got a little nerdy this weekend.
I took some existing dyno numbers from Drift Office that were available that had a baseline for the stock BRZ and a JDL UEL header
which you can find here.
From there, I extrapolated gains for the new 4-2-1 EL header vs stock. Since the 4-2-1 was compared against the JDL UEL header on the dyno from JDL; I could get a rough idea of the gain over stock from the Drift Office dyno using percentage gain of the 4-2-1 vs the UEL on JDL's dyno numbers at each 100RPM datapoint and then applying those values to the Drift Office JDL UEL numbers.
Keep in mind that this is very high level and does not include differences in tuning, etc.
Interactive version
Interactive version