[...] it has no memory. In case you charge a small battery that is capable of 0.25A charging current and power supply cuts off and then recovers, the charger defaults to 1A charging. In worst case scenario you may get fireworks....
That's a fairly serious design flaw. If this can occur while charging Li-ion cells then that could prove to be quite dangerous, since charging Li-ion at too high current is one of the primary causes of thermal runaway.
It would be very helpful if reviewers tested charger behavior in exceptional cases like this. Typically reviewers do not perform such testing. Yet knowing how chargers handle such exceptions is essential knowledge required in order to properly assess their performance across the entire safety spectrum.
This is especially a concern for chargers like this that are designed to make automatic choices (for novices, or for convenience, etc). Often, as above, they do not have access to information required to make decisions, so they are forced to choose between making very conservative (but inconvenient) safe, default choices, vs. agressive (more convenient), but possibly dangerous default choices (e.g. default charge current). And, what's worse, in some cases the designer may not have even considered some exceptional conditions, which increases the chance that unsafe (default) choices might be made by the firmware.
With competent design you get a plug-n-play charger. But with incompetent design you get a plug-n-pray charger. To accurately distinguish the two requires much more extensive reviews that test performance in cases like these.
This is an area where everyone can contribute,since many such tests do not require any special knowledge or equipment. Rather, it requires only that we devise some standard safety test [procedures (tweaked as need be based on the features of each chargers).