As an engineer that runs my own business and has to market - is designing a component of a product, that my customers replace on average every 10 years, with an eight year service interval REALLY good engineering? You can cut materiel/quality costs and reduce service interval to five years - and deal with irate customers that don't service and have components fail - their fault, but bad PR. Or you can increase materiel/quality to have a 10 year service interval - in most people's eyes maintenance free. In which case why not save the money required to make it serviceable (usually peanuts, but at high volumes....)

Had several interesting discussions about this.