The subsurface has generally a 3-D structure. The MT method is unique in that the impedance tensor and the tipper vector inform us about subsurface dimensionality. This implies that it is possible to know if 1-D or 2-D inversion is a reasonable approach or if a full 3-D inversion is needed. Simply speaking, 1-D inversion may be enough in sedimentary regions in which one is interested in slowly varying thicknesses of aquifers and aquitards, while 2-D inversion is suitable in which there is a clearly defined 2-D structure in tomography and geometry. By measuring a profile perpendicular to this "strike" it is possible to infer information about steeply dipping contacts between aquifers and aquitards. Without a careful dimensionality analysis, the best one can do is to ask:
Is the Earth likely to be explained by a number of continuous layers? 1-D might be ok. Is there one spatial direction in which geology is not expected to vary? 2-D might be ok. Is there strong geological variations in three dimensions? 3-D is most likely needed.
Ten years ago, 3-D inversion was really at the forefront, today it is becoming more widely available and easier to use. One important complication with MT data is related to so-called "static shifts" caused by smaller-scale heterogeneity in the near-surface. The use of the MT method is only recommended after proper training or study (e.g., using the book reference below).
Reference: Chave, A. D. and A. G. Jones, 2012, The magnetotelluric method: Theory and practice, Cambridge University Press.