Even if you model it correctly, you will need to place it at one of the multiple quarter wavelength points along the exhaust. These points change with velocity and temperature so your power benefits will alter in rpm slightly depending on temp and flow, which in turn also depend on increased power out and then this causes additional temp rise.
Your model needs to include accurate flow modelling for WOT and at full racing temps if it is for the track.
One issue with 90 degree placement of helmholtz resonators is the flow itself can disrupt the end condition for the resonator neck and prevent it working well. For example blow across a bottle to make it resonate and then blow very hard, it often changes. Or use a penny whistle, gentle blowing over the holes produces the fundamental tone, a harder blow increases the pitch from 1/12 wavelengths (odd harmonics) so at the frequency desired, you get no effect but it could help at 15000 rpm lol.
You can minimise this effect with a fine mesh over the neck as it makes it more consistent but likely to be compromised overall. If you are modelling in CFD, you also should use acoustic modelling which is sympathetic to changes in flow or your model will always be inaccurate and bench testing is inevitable.
Good luck, it's complicated!