I don't think it's a serious study to count the fire incidents in a single year (2023) and then use it as a generally valid statistic for an entire product cate

I don't think it's a serious study to count the fire incidents in a single year (2023) and then use it as a generally valid statistic for an entire product category. Firstly, the interval is too short, whereas other product categories have many years, if not decades, of statistics. For low probability events, counting in too short a time frame can significantly distort the statistics. Secondly, household battery systems are all relatively new. Similarly, EVs are a recent phenomenon. Is it not reasonable to expect that the chemicals, and hence the safety, of these batteries will degrade over time, so that the fire hazard will increase over time? And if so, counting incidents only for newly installed systems is obviously biased. And third, are all household storage systems the same in general design and chemical composition? If not, then lumping them together is unlikely to be very helpful in understanding the relative safety of the product category.