Critique the survey methodology if you think it is inadequate. If you are just questioning the whole concept of polling then it seems sort of willful ignorance.
Not that it assures a representative sample, but the poll was done by Ipsos, one of the larger global market and public opinion research firms.
Also at 90% you have a fairly high assurance of what the general consensus is.
Only assuming that the sample is actually representative. If I just polled 10 people in San Francisco on something, a result of 90% doesn't mean you have anything resembling a general consensus
There might be sections of Peruvian society IPSOS can't reach, but I don't think their methodology is that flawed
It might not be flawed, it might just be deliberately biased. But who knows, I certainly don't.