On the relevance of the multiplicative anomaly