Well, a collision or eight is inevitable. Every single hash algorithm that has been and ever will be will have this problem. A summary of a fixed size based on something bigger, without size restrictment, will always have an infinite number of collisions.
The question is, what sort of a collision are we talking about. As far as I know, the SHA1 algorithm is pretty decent when it comes to prefix/suffix collisions, as there are yet to be any collisions of these types. A random collisions between two completely unrelated inputs wouldn't really be a significant problem, because the inputs are unrelated. However, a prefix or suffix collision is a problem because it can potentially shorten the amount of needed attempts in a brute force attack.
What would be much worse is if the SHA1 algorithm could be shortened mathematically. But that is a totally different story.
