Of course they also tried a 'Frangible' hatch as well ...this copied from 'Flight' 1956,but this also was not without problems and I do not think lasted very long (cannot remember why)...
Canberra Navigators' Ejection
To improve the emergency escape facilities of the English Elec-
tric Canberra B.2, a frangible Fibreglass hatch to replace the
metal unit now fitted over the navigator's compartment has been
designed by the Martin-Baker Aircraft Co., Ltd. At the same
time the company has developed from the existing Canberra seats
a new fully-automatic ejection seat fitted with a duplex drogue,
an 83ft/sec telescopic ejection-gun, arm-rests, and a secondary
firing handle on the front of the seat pan. Thus equipped, the
navigator should be able to fire himself through the frangible
hatch and descend safely from almost any altitude. Canberras at
present have the Mk 1 non-automatic seat, and parachutes carry
barometric rip-cord release devices.
It has been found by experience that, without the explosive
hood-jettisoning gear linked to the seat-ejection mechanism, it is
preferable to eject through a canopy rather than to jettison it
first. The crew member is thus protected from air blast while he is
reaching for the firing handle, and a number of pre-ejection actions,
with their associated delays, are eliminated. When g forces are
operating on the aircraft, as may well be the case in an emergency,,
any pre-ejection action may become excessively difficult andv
precious time be lost. It is not, in any case, necessary for a good
ejection that the canopy should come away cleanly.
Following a series of successful test ejections, on static rigs and
in the air, a demonstration in the presence of Service representatives
was given at Chalgrove airfield on January 19th. A
"standard dummy man," using the fully-automatic seat described
above, was ejected through the new frangible hatch from a Service
Canberra flying at about 300 kt at 400ft. The aircraft was piloted
by Capt. J. E. D. Scott, the company's chief test pilot. The test is
stated to have been successful in showing the practicability of
ejection through the new Fibreglass navigator's hatch, and also
in demonstrating the advantages of the procedure. Both the Fibreglass
hatch and the new ejection seat have been designed to be
easily and economically fitted as a retrospective modification