Define bad though? Movement regulation and adjustment tolerance has moved on massively over the years. i remember when micro brand watches with Miyota movements or Seiko NH35s were running either side of 30-45s a day. That was pretty poor but that was what it was like in the 2012~ era.
A Casio digital will be accurate to usually 1s a month, you can never expect that level of timekeeping from a mechanical movement just left in one position for a week for example as it will gain or lose consistently at a fixed level (assuming it is kept wound then put back). But if worn and stored with a routine and if the movement has a level of adjustment/regulation with multiple positions then there's no reason that after a month the movement isn't not far off similar accuracy. Like for example one day it will gain 3s, the next day it will lose 2s, the following day it gains 4s, the day after it loses a second and so on, so whilst the daily variations will be such depending on how you wear it, or how it's stored and level of power reserve, the overall accuracy over the course of a month will mean it is usually closer to 0 anyway.
The problem only arises on poor quality movements or magnetised movements that are in the +30s a day range which is pretty common especially with Seiko movements. Unless it's an Ali Express special etc, then I do not expect to see any mechanical watch bought today to have poor accuracy really from the majority of micro brands or bigger names that aren't the big Swiss/German names.
There are different qualities and grades that will dictate how accurate they are too. Generally, most brands whether first party of micro will to in-house adjustment and regulation anyway though and you can typically expect within +/- 10s a day. The vast majority will be within 0-4s a day, especially the higher grade movements.
