Good engineering, and good mechanics, have the principle of least change: make the smallest change needed to achieve the desired outcome.
The principle applies to whole bicycles. Beyond a certain age, beyond a certain level of disrepair, do little.
Great old frames deserve rebuilding — if the customer has the budget.
But don’t over-write and overwork a beat-up bike. In the end, the customer could have better spent the money. In the end, the bike will take more work then charged.
Old bikes have unpleasant surprises: bearings that seemed dry turnout blown; cables reveal rust; etc.
Do the least possible — or do everything, and charge for it. Anything in between rarely works.