Hi wellshao,
A proper answer to your question requires a little more details about your attempt at "mapping" your phenotype (your experimental design and/or methodology). Yet, my guess is that he is trying to tell you that the experiment you ran to analyze your transgenic strain could be giving you misleading results.
Let me give you an example: let's say you are looking at the expression of a given gene in your strain. You use standard RT-PCR and run a gel to see how much this gene is being expressed in a wild type control compared to your transgenic strain. Based on your gel you determine that there is no difference in expression since both bands have the same intensity. Yet, because this type of PCR only looks at the end product (what accumulated after 40 cycles of PCR), differences that occurred during PCR (from cycle 1 through cycle 40) may well be hidden. For example, if the amount of PCR product reached maximum accumulation at cycle 30 for your strain and maximum accumulation at cycle 39 in your wild type control, you would not be able to see this difference by running these reaction on a gel since at cycle 40 both would show the same band intensities. Yet, if you were to use real time RT-PCR, which can detect DNA accumulation at every cycle of PCR, you would be able to see that your strain reached maximum accumulation at cycle 30 and the wild type control reached maximum accumulation at cycle 39, which means that your gene of interest is expressed by as much as one thousand fold more in your strain compared to the wild type control. This is an example of why "kinetics are essential since differences in outcome may not be absolute".
Hope this makes sense. It depends a lot on your experimental design.