You can trust the meter once you know what and how it is reading light, and what to do about differences.
Digital cameras, among so many others (including film cameras), use multi-pattern/matrix/evaluative and calibrated to a scale of optimal exposure for the method (digital) — at any time these meters are reading a great amount from the scene, more often than not with consummate precision. But making a reading with a digital camera and then to expect a hand-held meter to match the reading, well, the 'how and why' is because of the method being used. An simple incident reading is not the same as what the camera is reading, which is looking at several elements (light tones, dark tones, background light, spectral light, hot spots etc.) of your set up to strike an effective balance, as opposed to an overall averaged incident reading you took with the Sekonic.