The hypothesis that the blue stragglers in the low-density globular cluster NGC 5053 are merged stars formed as a result of physical stellar collisions during dynamical interactions involving primordial binary stars has been investigated in detail. This was done by carrying out a large number of binary-binary and binary-single scattering experiments in which the stars were allowed to instantly merge if they touched. Uniform distributions of binary periods P in log P space between various limits were considered. It has been found that collisions between binaries with periods exceeding 10(2) days are the best at producing merged stars that remain bound to NGC 5053. The distribution of binary mass ratios is much less important than the distribution of periods in determining the rate at which stars collide. Binary-binary interactions are almost an order of magnitude more effective than binary-single collisions at producing mergers. In the limiting case of a binary frequency of 100% in the core of NGC 5053, we find that, depending on the period range considered, 28%-55% of the cluster's blue stragglers can be accounted for by physical stellar collisions during binary-binary interactions. With the binary frequency in the core reduced to 50%, binary-binary and binary-single collisions can still account for 9%-18% of the blue stragglers. These percentages are roughly doubled if one includes tidal capture binaries formed during dynamical interactions involving binary stars, since these close binaries will produce blue stragglers via gradual coalescence or mass transfer. Given the fact that a surprisingly large number of primordial binaries have been discovered in NGC 5053 by Pryor and his collaborators, it is difficult to reject the possibility that at least some of the cluster's blue stragglers have a collisional origin. The characteristics (e.g., degree of central concentration, mass function, binary properties, and rotational rates) of the merged stars are described.