Consider decreasing the confidence threshold of Tesseract

[laurent22]

It seems most apps that include OCR can detect the text below. We should decrease the confidence threshold so as to make this particular case pass.

Ref: https://discourse.joplinapp.org/t/still-some-way-to-go-for-ocr/42964/3