On the influence of rain gauge performance on extreme events statistics: the case of weighing gauges