The key point, as I'd understand it, is this: nothing can completely eradicate the risk of contamination - this is known and understood in justice systems worldwide. But it's also generally accepted that if law enforcement and forensic science personnel carefully follow established guidelines with regard to evidence collection, handling, transportation, chain-of-custody, testing and interpretation, then contamination can be (to all intents) excluded - unless the defence can actually prove contamination.
But the flip side to this is that if correct procedures and protocols are not followed, the burden of proof regarding contamination progressively shifts towards the prosecution having to try to show that contamination could not have happened in spite of the breach of protocols/procedures.
There's a very, very good reason, in other words, why national and international standards/procedures/protocols are developed and agreed. And once they've been agreed and implemented, they're not simply a "nice to have" option - they must be rigorously adhered to (unless there happens to be a very good reason not to do so).
A micro-example of this might be the area of low-copy range DNA testing in this case. Some people seem to think that the argument begins and ends with the binary question of whether LCN DNA evidence is admissible in and of itself. These people don't understand that its admissibility - if it's admissible at all - is entirely contingent upon extremely rigorous standards of cleanliness and air handling in the testing lab, and multiple amplification/testing/interpretation of the low copy sample. I think I'd be correct to say that every court in the developed world which has accepted LCN DNA evidence - with the exception of the court in the first trial of Knox and Sollecito - has only done so if the LCN analysis has been conducted under these extremely strict protocols.