In my humble opinion, luciferase reporter mice are better exploited to track responsive elements of critical transcription factor, in that case the photon emission would be a surrogate marker of the final result of one/more signalling cascades and not just the mirror of one of our 30,000 genes. That's much informative. Of course, then you need some deconvolution, but you can reasonably get the dissection of your pathway by crossing your reporter mouse into a KO background. Definitively, to study mRNA expression in a tissue, I prefer a hierarchical clustering of quantitative PCR dataset (of different genes), chiefly with new technologies (here and here). Yes, this is just my personal naive opinion, feel free to fill my gaps.
K. Birsoy, A. Soukas, J. Torrens, G. Ceccarini, J. Montez, M. Maffei, P. Cohen, G. Fayzikhodjaeva, A. Viale, N. D. Socci, J. M. Friedman (2008). Cellular program controlling the recovery of adipose tissue mass: An in vivo imaging approach Proceedings of the National Academy of Sciences, 105 (35), 12985-12990 DOI: 10.1073/pnas.0805621105
this is another example of BLI crack down