As per my knowledge, it's one of the best method to understand the behaviour of birds in its natural habitat rather than its Biodiversity 

The short answer is yes. There is a burgeoning literature on this. The big question is how does one do so most efficiently. Some papers in a Special Feature of Avian Conservation and Ecology cover this, and I know of several in review and in preparation that also address efficiency. The Special Feature of Avian Conservation and Ecology is freely accessible here: 

https://www.ace-eco.org/issues/view.php?sf=5

Yes! :)

Not only possible but necessary, in addition to other methods

Here you can find a project that is using sound recorders to monitor the vocal behavior and diversity of birds in forests. It is true that in that case most of the studies were at species-specific scale.

http://cobra.ic.ufmt.br/

Here, it is a recent link from where you could start to read (including the references therein) a little more on the topic. It is a "feasible" way of working at community level avoiding the need of looking for every bird species in the recordings.

Article Rapid assessment of avian species richness and abundance usi...

Multi-species occupancy models (Iknayan et al. 2014 Trends in Ecology and Evolution, Furnas 2020 Biological Conservation) are a great way to get the most out of community-level data from ARUs. There is stronger inference about common species, but data augmentation allows some inference about rarer species and estimation of total species richness including those rarest species that were never detected.

Darras, K., Batáry, P., Furnas, B., Celis‐Murillo, A., Van Wilgenburg, S. L., Mulyani, Y. A., & Tscharntke, T. (2018). Comparing the sampling performance of sound recorders versus point counts in bird surveys: A meta‐analysis. Journal of Applied ecology, 55(6), 2575-2586.