Hi,
Staking is done to heavily milled TR race heads to prevent the thinned head collapsing into the cooling passages.
With the head removed, you should be able to very easily tell if there is any problemwith it. Lay it face up and put a straight edge across the head to see if there is any warpage and look closely for any damage or collapse of the sealing surfaces right around where the rim of each cylinder, where it contacts the head.
The original TR 4-cyl. heads are nice and strong, until they're milled a good deal to significantly increase compression. TR2/3 heads need a little more work to be milled as much as later heads can be, but the primary trouble area with the earlier head is interference between the water pump and the thermostat neck. Still, the earlier heads cannot be milled as much as the later ones, because there isn't as much metal cast around the compression area.
However, unless you have .120"+ milled off and 11:1 or higher compression going on (do you need to run 100+ octane racing fuel?), staking is unlikely to be needed or helpful.
With a street engine that's been milled less (say .030 to .090") and is under 10:1 CR approx., I think it's far more likely that there's a problem with the cylinder liners not standing above the face of the block enough. There should be .003 to .005" protrusion to effect a good seal with the gasket... and this is adjusted by using thicker or thinner figure 8 gaskets under the sleeves, or by skimming the end of the cyl. sleeve or skimming the top of the block slightly, depending upon whether more or less protrusion is needed.
By the way, while it's off the block is a good time to measure the head and estimate how much it's been milled, if you aren't certain. Originally, I think it should have been 3.30" thick (someone correct me if I'm wrong... this number might be 3.330"), from the sealing face to the top flange. Get some good measurements at various places around the edges to compare with this number and try to determine about how much your car's head has been milled, if at all.
Yes, as mentioned, another good possibility is improper torque. Old stretched or thread-damaged head studs or nuts are often the culprit, causing improper torque by giving a false impression that enough torque has been applied (when the threads bind). Also flares or belling on the threaded stud holes in the top the block can hold the head off from making a good seal, in spite of "proper" torquing.
I hope this helps. Keep us posted what you find.
/ubbthreads/images/graemlins/cheers.gif