Are you making these observations with housing portions that have a
cable with tension on it inside? And with housing portions neatly
squared off to butt cleanly as has been suggested ?
What force is it that you think will make the housing sections slip
laterally? (it isn't coming from cable tension) If they slip laterally
so easily, how are they going to apply sufficient sideways force (so
as to cause abrasive wear to the cable?)
Sure you can "force" lateral displacement of the housing portions but
normal cable tensions are not going to cause it. And the tape will
maintain alignment against any minimal lateral forces. In fact, in
anything other than a perfectly straight run, the cable will tend to
lie against one side of the housing and keep the housing portions
aligned.
How many layers of tape are you using? More than one is probably
useful.
I don't think anyone has suggested that a cleanly butted and taped
junction will match intact housing in bending resistance, but you have
hypothesized an uncommon situation. And barring a kink of the inner
cable, once tension is reapplied the housing pieces will re-butt and
work essentially as originally configured.
Tape (or heat shrink tubing) has the advantage over metal tubing of
more positively sealing the housing against moisture or other
contaminants. But that could be reinforced with metal tubing if
desired. Small brass and aluminum tubing is readily available from
"hobbyist" sources. I know that the Ace hardware store here carries
this line:
http://www.ksmetals.com/products.html
While it is perfectly feasible, you needn't mate cable housings if you
choose not to.
DR