There are a lot of different figures on tube life and heater voltage, in part because different tubes respond differently. Those with thoriated heaters, for example, do well when run at rated voltage for their first 200 hours, then dropped to 95% of full voltage until they approach end of life.
http://www.cpii.com/docs/related/22/AB18.pdf Dropping to 90% of rated voltage after the initial 200 hours may increase filament life to the point that heater failure will be eliminated as a cause of (thoriated) tube failure.
http://www.burle.com/cgi-bin/byteserver.pl/pdf/tp202.pdf Tubes with oxide-based heaters are different. Power tubes can have their life shortened by running at lower heater voltages, but others often benefit from slightly lower voltages: at 10% above rated voltage, life will be cut by half, at 5%-10% below rated voltage, tube life will be significantly enhanced at the cost of slightly worse performance in various other regards.
http://www.pmillett.com/Books/Tomer_1960_Getting_the_Most_Out_of_Vacuum_Tubes.pdf I suspect that the author cited by the earlier poster was thinking of power tubes when he condemned running tubes at reduced heater voltages, since it is very true with those.
My recommendation would be to never run power tubes at anything but their rated voltage, and to rectify, filter, and regulate that voltage if you're using expensive tubes. For other types of tubes, if high longevity matters more to you than squeezing every last bit of performance out of it, rectification and filtering are also good, but the target voltage should be 5%-10% below rated value. Anything lower than 90% of rated value will cause worse performance without enhancing tube life, and may actually shorten life. Using tube coolers and/or cooling fans will also increase life, often doubling it.
Using 9v in place of 12.6 for heaters is not a great idea from any perspective, in this schematic it has to be considered a matter of convenience rather than good design.
