I'll start from the second point. The usual criteria to "locate" an endangered language are:
- The number of native speakers that are still alive (at the moment of the analysis);
- the average age of these native speakers (or fluent speakers);
- the percentage of the new generations that acquire fluency in that given language.
If a language stops having native speakers, but it's still used in some contexts, it becomes a dead language, such as Latin.
The number of native speakers can be a hint, but it doesn't necessarily mean that a certain language is disappearing. For example, a language can have millions of speakers, but if nobody learns it as a child, or if the speakers will "move" to the national language abandoning the local varieties — it seems this is happening in Indonesia — then those local languages will be in danger. If a language has some hundreds of speakers, but it's the primary or only language of a certain population, it won't be in danger.
The Ainu language is considered to be in a dangerous situation, having only 300 native speakers where only 15 use the language on a daily basis. Not to mention that the replacement by the new generations is not sufficient.
Also the Leonese language appears in the list, and many others, that you can find in this article from the Guardian, that features a list provided by the UNESCO. The UNESCO also provides an Atlas of the World's Languages in danger.
Some linguists think we should save these languages, because they belong to human history and in turn can help to decipher it and understand it. Other linguists think we should push towards a world situation with less languages in order to favor a good communication between people and possibly arrive to a single global language.