Colour can play a part in design and it plays a big part in making something look 'good'.
However, this article makes some misleading points, and it isn't all as it seems.
Result 1: You can't convey smell or texture online so it's irrelevant when it comes to e-commerce. Plus, it really depends entirely on the product, so the statistic is fairly meaningless.
Result 2: A 'particular product'? The only thing those results could remotely apply to are clothing, pens, buttons, ribbon, paper or anything else where colour is one of the few major variations of a product. For everything else, that statistic is utterly wrong in my opinion.
Result 3: 'Colour increases brand recognition by 80%'. Compared to what? Black and white? I cannot see how they could accurately measure this anyway.
'Colour is one of the most powerful methods of design'. Colour is a method of design?
'Colour is not the only item that influences customer behaviour'. Really

.
'Percentage of shoppers who base their opinion of a website on overall design alone'. I find it very, very, very hard to believe that 42 out of every 100 people manage to blank out every element of a website other than design to form an opinion. What people say, and what people do, are often very different things.
'52% of customers are more likely to enter a store if there is a sale sign in the window'. 'More likely' is utterly meaningless. How many of those 52% could still enter the store anyway even if they aren't 'more likely'? Plus, with some stores, putting up a 'sale sign' could actually do more harm than good.
The only things I can spot in here worth taking serious note of are:
- Speed up your website
- Use a good design
- If you're starting up a new website or new branding scheme, then it's worth looking into how colours can have a small influence.
Everything else could actually do some damage if people start re-designing their websites with magical psychology-busting colours and slapping up sale and guarantee signs everywhere.