Re: exacting precision
Technically, the former (in infrared wavelengths at least) is Raman spectroscopy, and the latter is transmission spectrography. There's no reason both can't be used together. True, occlusion is the only real way to determine if there is a planet there, but once you've determined tha, the variations in the Raman spectrum of the star/planet system as the planet orbits its parent's star, although small, could theoretically provide a diagnostic spectrum, and do it over a longer timescale than the short period when the planet is between the star and us (the observer), which could be a very short window of opportunity if the orbit is eccentric and not nicely aligned with us, or has a long period.
If it's sensitive enough, and you can separate the useful information from the background noise (selectivity), Raman spectroscopy could be used to detect atmospheres on multiple orbiting planets, as they move in and out of the range of view (to see the reflected light, they would need to be behind, or perpendicular to the parent star compared to the observer). I don't know enough about the sensitivity and selectivity of the method to answer this myself, but I probably know people who do/can.