Valid point... another issue is that of electrical design to help prevent unexpected ejection firing.
What I've done on all my "electric ejectors" is use a pair of rotary switches, mounted behind holes in the airframe so that their shafts can be turned from outside with a small tool. One is the master on/off switch for the altimeter. The other is a 3-position switch connected in series with the output circuit. When the switch is centered ("off"), the circuit is disconnected. (For extra safety here, one could refine the wiring so that the e-match is not only disconnected but also shunted, but I haven't done that so far.) One of the two side positions ("ready") connects the e-match directly, while the other ("test") has a 150-ohm current limiting resistor and a high-brightness LED in series with the firing circuit. (With a 9 volt battery, and 2 volts drop across the LED, this will limit the maximum possible current to about 47 mA: well under the rated "no fire" current for either "standard" or "sensitive" e-matches, and just within the 50 mA continuous current rating of the LED's I used.)
When I construct the circuit, I perform the following tests with a "bare" (no ejection powder) e-match of the type I intend to use:
First, with the switch in the "test" position, I verify that the altimeter is properly sensing "good continuity" through the resistor and LED. (In some cases, the continuity current is enough to visibly light the LED - if it's bright enough to cause potential confusion, I sometimes have added an additional resistor, selected by experiment, in parallel with the LED.) Then I try to "fire" the match - some altimeters have a "test" mode by which they can be manually made to close their output circuits; with others I simply use a jumper to complete the circuit across the output transistor.) If all is well, I see a *bright* glow from the LED but the match won't fire.
Next, I turn the switch from "test" to "ready" and repeat the test. This time the match should fire.
At pre-flight I assemble the rocket, and load it on the rod, with both switches off. First I turn on the power switch, and verify that the altimeter is giving the correct beeps/blinks/etc. to indicate "on but no continuity". Then I turn the output circuit switch to "test". The altimeter should now give a "good continuity" readout, but the "safety LED" should NOT light.
At this point I've assured myself that I have continuity in the output circuit as assembled, and also that the altimeter isn't trying to fire yet. This gives me confidence that I can now switch from "test" to "ready" without anything unexpected happening.
Obviously, this can't prevent all possible accidents, since it doesn't address events such as mechanical damage to (or a static discharge into) the wiring, so the general caution you mention is still applicable, but I feel that this circuit configuration will help to control for the most probable causes of unexpected firing.
-dave w