I suppose it comes down to what you want from the RT-qPCR. A one step RT-qPCR will allow you to process a larger number of samples, as it will be less hands-on, and will also reduce the chances of contamination or sample mix up. A two step will allow you greater flexibility in the amount of RT product added to the qPCR and flexibility in the priming you use for the RT step (oligo dT, randomers or gene specific priming).
From a personal point of view I preferred a one step RT-qPCR for the plant virus diagnostics I used to do as this reduced the handling steps and the chances of sample mix up or cross contamination.
I suppose it comes down to what you want from the RT-qPCR. A one step RT-qPCR will allow you to process a larger number of samples, as it will be less hands-on, and will also reduce the chances of contamination or sample mix up. A two step will allow you greater flexibility in the amount of RT product added to the qPCR and flexibility in the priming you use for the RT step (oligo dT, randomers or gene specific priming).
From a personal point of view I preferred a one step RT-qPCR for the plant virus diagnostics I used to do as this reduced the handling steps and the chances of sample mix up or cross contamination.
On balance, like Chris, I think they are largely interchangeable unless you are dealing with known low copy number genes; in which case 2 step can be optimised and made more sensitive
The only clear cut general preference I have, based on personal experience, is if you are dealing with small amounts of starting material (< 10^(5) cells); in which case combined RNA extraction and cDNA synthesis is without doubt better
But in terms of cDNA synthesis and especially PCR it is a matter of personal choice, available kits aptitude and experience etc