First, plug-in surge protectors would not even provide
effective protection meaning that they are not even relevant
to the originally posted question. Furthermore, the little
surge protector does not absorb even modest transients.
Absorbing is not what they do.

Second, a full-on lightning strike is why we install
protectors. Modems already have significant internal
protection as part of design and to meet industry
requirements. Protection that can be compromised by a full-on
lightning strike.

Typical frequency of potentially destructive surges is once
every eight years. That number varies significantly even
between adjacent towns. So how frequent is your
neighborhood? Without information such as underlying geology
and manmade buried objects, weather trends, etc; then your
only valid information comes from history provided by long
term neighbors. Yes, even installation of new buried
utilities can change those trends.

'Whole house' protectors and earthing is so inexpensive that
US telco companies install same, for free, at every customer
interface. Question is whether the £1 per protected appliance
is necessary for a destructive transient that might occur once
every ten or so years. Points one and two define why you
would install that protector.

Pyriform wrote:
w_tom wrote:
The bottom line is this. Nothing is going to stop
lightning. Lightning protection is about shunting - also
called diverting, redirecting, or electrically connecting - to
earth ground. Any protector that claims to stop or block
lightning (such as in the RCD question) is simply promoting a
myth. And yet that is exactly what many people do - promote
the myth - when they recommend plug-in protectors.

I am not naive enough to think that my little "surge protector"
would save anything from a full-on lightning strike. The issue
is whether one might reasonably expect more modest voltage
transients to occur on the mains supply (and phone line) which
can be safely absorbed by such a device, and which might
otherwise disrupt or destroy delicate electronics (equipped,
perhaps, with rather cheap PSUs).

I freely admit to not knowing the answer. Your contributions
make good points about the requirements to be met in order
to achieve 'proper' surge protection, but do nothing to
address this question.