You can use embedded Lucene if you want great search but no UI/schema management and you are happy to code the integration. You can start with 3 jars (5Mb) in size and add additional jars as your needs expand (e.g. multilingual processing).
You can use a single node Solr, if you want UI and send data in multiple formats and may eventually need to scale.
You can use Elasticsearch if you know you are starting big and deal with scaling from the start.
You have the choice. It is not just clustered Elasticsearch vs code your own.
well, I'm pretty good at all the non operations aspects of ElasticSearch, so in my experience if you have ES in some sort of managed cloud instance - like searchly, doesn't have to be AWS or elastic - it's a one man job - and also in my experience not necessarily full time devoted to just es.
May vary dependent on if you have a lot of data, but if so you probably have people to handle lots of data problems anyway.
You can use a single node Solr, if you want UI and send data in multiple formats and may eventually need to scale.
You can use Elasticsearch if you know you are starting big and deal with scaling from the start.
You have the choice. It is not just clustered Elasticsearch vs code your own.