The "10-degree hotter, half lifetime" rule in electronics is just the chemical reaction rate in the Arrhenius equation. It's fairly accurate for electrolytic capacitors, but probably not semiconductors. Even in the original 1970s MIL-STD document that made this rule the industry, it acknowledged junction temperate doesn't really matter below a point. It's just that there's really no good way to extrapolate lifetime besides testing it for 20 years, so lacking any alternative, people keep repeating this guideline...