That 85C is rise above ambient, so the actual temp of the filament bulb was 105C.
Yes. I always test all the lamps in a packet. I have been stung in the past by inaccessible bulb locations!
At work in the 1970s we had big numeric counter displays where each pane carrying an engraved number was top and bottom edge-lit by four 24V bulbs, 0-9 and times eight. This meant 300+ fiddly midget-flange bulbs to replace all at once on a maintenance day when they began to fail. We had several of those counter boxes.
So a little test jig was made and each new bulb dabbed in and if the light meter reading was in the correct zone then the bulb was OK. Testing like this took just a second or two, and was justified by about 5% of new bulbs not up to spec or dead.
More recently (2000) I did something like this at first with ultra-bright green LEDs for a project (108 at a time on a PCB) until we could persuade the manufacturer to grade them by intensity and colour. Our effort was more expensive than theirs.
And just now the Noobibaba were within 5% of each other in brightness on my old Weston light meter and about 10% more than the brand new filament. 
(I guess about 5% experimental error)