ElasticSearch Query Score Precision

Elastic Search
1

Hello,
I am planning to build a Rule-Matching engine over Elasticsearch. Requirement is to assign exponential weights to each parameter. For instance, if there are 30 parameters, param1 will have weight equivalent to 21, param2 -> 22 .. param30 -> 230.

Query to Elasticsearch would look like below

{
  "query": {
    "function_score": {
      "query": {
        "bool": {
          "must": [
            {
              "bool": {
                "should": [
                  { "term": { "parameterA": "valA" } },
                  { "bool": { "must_not": { "exists": { "field": "parameterA" } } } }
                ]
              }
            },
            {
              "bool": {
                "should": [
                  { "term": { "parameterB": "valB" } },
                  { "bool": { "must_not": { "exists": { "field": "parameterB" } } } }
                ]
              }
            },
            {
              "bool": {
                "should": [
                  { "term": { "parameterC": "valC" } },
                  { "bool": { "must_not": { "exists": { "field": "parameterC" } } } }
                ]
              }
            }
          ]
        }
      },
      "functions": [
        {
          "filter": { "term": { "parameterA": "valA" } },
          "weight": 8
        },
        {
          "filter": { "term": { "parameterB": "valB" } },
          "weight": 4
        },
        {
          "filter": { "term": { "parameterC": "valC" } },
          "weight": 2
        }
      ],
      "score_mode": "sum",
      "boost_mode": "replace"
    }
  },
  "size": 1
}

How precise would the scores computed by Elasticsearch be?
Will the weight of the lowest parameter be nullified, when some parameter of higher weight (say 230) matches?

The total number of matched records will be less than 10k.

I can think of two approaches to solve this usecase:

  1. Leverage Elasticsearch's weightage/scoring functionality to directly fetch the record with the highest score.
    Concerns: Score Precision, High resource usage in Elasticsearch
  2. Fetch all the matching records in application, and compute the scores inside Application.
    Concerns: High Latency & Network bandwidth to fetch all the matching records, High resources usage in Application to compute scores

Thanks