The normal MOV failure mode is not to open, fail shorted, or
vaporize. Catastrophic failure occurs when the MOV is
installed as not intended by its manufacturer. A condition
known as catastrophic failure due to a grossly undersized
protector. Manufacturer provides life expectancy data
sheets. MOVs degrade according to number and strength of
transient current.

However, when selling ineffective MOV protectors to a naive
public, some companies undersized them. Then when a grossly
undersized protector fails (and when protection inside the
appliance actually protects the appliance), then the human
says, "The protector sacrificed itself to save my computer."

Effective protectors degrade. Grossly undersized protectors
are undersized to be sold at excessive profits; so that a
human will recommend it and buy more. Take a $3 power strip
from Home Depot or Wall Mart. Install some $0.10 components.
Sell the modified $3 power strip for $15 or $50 as a surge
protector. A naive public will suffer catastrophic protector
failure - a completely useless protector - then buy more and
recommend the overpriced and undersized device to friends.

Protectors might be placed either on the input or output of
the conditioner. On the input, a shunt mode protector would
simply provide a destructive transient with more wires into or
around the conditioner. If the conditioner is working as
intended, then a power strip protector connected to
conditioner output would do nothing. The 120 volt power strip
protector should ignore everything until voltages exceed the
let-through voltage (about 300 volts). "Should" because power
conditioners may not stop or eliminate those typically
destructive 300+ volt transients.

If the power strip is connected to a typical computer grade
UPS, well, those UPSes output excessive voltage spikes when in
battery backup mode. For example, this 120 volt UPS outputs
two 200 volt square waves with up to a 270 volt spike between
those square waves. Spikes that do not harm computers. But
spikes may degrade a power strip protector or cause
conditioner component failure. (Note the word degrade; not
MOVs that short, open, or vaporize.) Protectors are not
intended for protection from repetitive spikes as demonstrated
by MOV manufacturer's chart described in 1st paragraph.

Is this UPS output destructive? Yes to some small electric
motors and power strip protectors. But internal computer
protection would make that 200+ volt spike completely
irrelevant. It's 'dirty' electrical output during battery
backup mode is why power strip protectors should not be placed
on UPS outputs.

Last few pictures demonstrates how to remove MOVs from a
power strip. Note the caption. MOVs are removed and the 'OK'
light still says the power strip is OK:
http://www.zerosurge.com/HTML/movs.html

Mike Rivers wrote:
I suspect that one reason why they might want you to hang a surge
suppressor on it is because when they're suppressing, they shunt
current that might otherwise be drawn by the connected equipment when
it gets more than its rated working voltage. This would put an
excessive load on the UPS/Conditioner. Also, although the general
failure mode of a MOV surge supressor is open (giving no protection
once it's failed) they've been known to fail in the shorted (or more
correctly, low resistance) state, which would put undue load on the
source.

I don't think it has to do with chaining anything, nor do I see a
problem with an external circuit breaker in line with the load. If you
want to use an outlet strip, take it apart (which might require a
drill or file on an outlet strip that's too cheap to be held together
with screws) and clip out the MOV. It's a round thing with two leads,
about the diameter of a nickel.