convert_datatype could have improved coerce flexibility

[geekpete]

For all general issues, please provide the following details for fast resolution:

• Version: 6.6.0
• Operating System: Ubuntu 18.04.2 LTS
• Sample Data:

This was the farequote dataset.
CSV file with a field that contains numeric value that contains a comma, eg 10,144.183

"February 11th 2017, 18:59:25.000",AVjeRNz82UMJWf0OwkY7,"farequote-2017",,responsetime,JBU,"10,144.183"

• Steps to Reproduce:

Found that Dissect's convert_datatype cannot coerce the number with the comma
for "thousands", resulting in this warning:

[2019-02-08T15:01:06,484][WARN ][org.logstash.dissect.Dissector] Dissector datatype conversion, value cannot be coerced, field: responsetime, value: 10,144.183

Whereas the Mutate filter's convert was able to coerce it.
This means having to use an additional mutate filter block to handle the conversion to get
this data into a numeric type.

I also saw this especially illuminating ticket: #10

which goes into some of the detail of how different the implementations are between this plugin and others for seemingly similar functionalities, so will understand if this enhancement isn't a trivial ask.