Thanks, Bill. They mostly seem to affirm what Gigabyte is claiming, i.e. solid capacitors are the way to go. Except for your first link, where Stele appears to be talking through his ass/arse: Solids are not necessarily better than 'wet' electros at high temperatures - the improvement comes as temperatures fall, but is good enough at 65C to be considered a major advantage.
To be more precise, life expectancy of conventional electrolytics doubles for every 10C drop in temperature, but it more or less trebles with 'solid' capacitors.
I'm also curious how he figures the solid polymer can still leak - from what I can glean, there are no visible indications when a capacitor has failed ...