• Following is the link for the DNA barcoding system with which you can pool several samples. http://www.biooscientific.com/next-gen-sequencing/ion-pgm-adapters-for-dna-seq/nextflex-dna-barcodes
• The size of target is not the first thing to consider, the read length output is 400bp for Ion torrent systems, so the max read length you can get will be 400bp. So the length of your target does not matter much.
• Different type of target might effect the sequencing efficiency but principally it does not matter what you sequence.

Abhijeet Singh thanks for your answer, but may be i wasn't clear enough to describe my question. I'm asking about the pooling step itself, as i have several barcoded libraries related to different applications.

Each application requires different amount of data (reads/bases) finally in order to be analysed with optimum depth of coverage, so pooling step should represent this variation after normalizing all the libraries to the same concentration.

Indeed it was not clear. It would have been better if you also have mentioned that you need different coverage of different libraries.

I think you can calculate based on the ION torrent sequencing platform you are using and the type of libraries you have with the formula

coverage = (Number of amplicons x length of amplicon)/total data generated