1764 people answer the poll, when the number of registered voters in France is just shy of 50 millions

That in itself doesn't invalidate the poll. There are standard statistical methods for figuring out how many people you need to ask to get a representative sample and it can be a lot less than one might imagine. But whether people were really chosen in an unbiased way is another question. A good survey will publish its methodology.