Updated passbands are different from previously released versions

[Raindogjones]

There are small but non-negligible differences in the passband tabulated intensities calculated using the feature-blending branch (and currently hosted in staging) c.f. those that were previously shipped.

I suspect that this is due to a change in the way we integrate intensities across the bandpass - the previous implementation was the np.dot of the bandpass and the intensities, while in feature-blending we have switched to np.trapz