[turretslug]

Is it easy to try it and see?- or would it involve changing the valve socket for a skirted type? I suspect that the best way to minimise temperature rise in a valve is to leave the envelope exposed- plain, i.e. bare metal cans were criticised for keeping heat in and blackened outer reduced the problem rather than being better than no can. The ex-professional use kit that abounded in skirted sockets and screening cans probably did so for retention and screening reasons rather than cooling- it's noticeable that hot-running valves such as rectifiers and small PA valves tended to have spring or fabric string retainers even when other valves were canned.

Also, valves like metal octal and many B9a pentodes are screened anyway, but not usually B7g types. As an empirical example, the B7g EF91 LO in my Eddystone 750 has a plain metal screening can as standard (this valve having an unscreened anode as outermost element), lifting the can changes frequency slightly (and unsurprisingly) as it represents stray capacitance to anode, I would expect a B9a type with outer mesh screen to be less affected. Also, spraying the originally plain can with a single thin coat of matt black paint made a very marked difference to warm-up drift.

Probably hen's teeth now, but some expensive kit like Marconi's TF144H had oscillator valve fitted with rather posh, sturdy cans that had an inner tensioned copper braid girdle that pushed snugly round the valve, the can locked down to the socket with several turns of a fine brass thread- I expect this type of screening can genuinely would keep the valve cooler than being open to the air.