In this paper, I read that mutant versions of Cas proteins such as a deactivated Cas9 (dCas9) are used alongside a guide-RNA (gRNA) to form variants of CRISPR tool that can function as transcription regulators.
Since the mutant Cas9 has lost its nuclease functionality, the mutant gRNA-dCas9 complex can "block" transcription factors such as RNA polymerase from binding a certain gene, hence repressing transcription or elongation.
Another way that dCas9 can regulate transcription is by "carrying" another transcription factor - be it an activator or repressor - and this can be done by engineering a dCas9 fused to the other transcription factor of choice.
I understand how the dCas9 can block the binding of another transcription factor to repress transcription initiation or elongation, but what is the point of fusing another transcription factor with the dCas9? It seems to me that in this scenario, the dCas9 is just a passive carrier of the transcription factor that is actually doing the job, so why can't the transcription factor be fused with the gRNA instead, so that it can directly affect transcription when the gRNA forms complimentary base pairs with the target gene, instead of it being on a passive dCas9?