The clue is in the name of the poll company. Lord Ashcroft Polls. Common sense dictates that you need to look at who commissioned a poll to detect if there may be any bias. Would you expect a survey of breakfast cereals commissioned by Kellogg to conclude that Nestle was best? So, yes, the points do have a bearing.

The number surveyed is not as relevant as how they were surveyed. Do we know how it was done? Random sampling? Active sampling of a larger base? Internet survey? Telephone? A combination?

I think we should stick to the larger pollsters like Mori and YouGov who are transparent in their methods.