I claim that there are no bad sequences in C(&omega;<sub>1</sub>).

Suppose to the contrary that x<sub>&alpha;</sub> is bad. For any countable
ordinal &beta;, there is r<sub>&beta;</sub> in
C(&omega;<sub>1</sub>) such that the distance between r<sub>&beta;</sub> and
x<sub>&alpha;</sub> for &alpha; &lt; &beta; is at most 1.
For any countable limit ordinal &beta; and any positive
rational number &epsilon;, there is a smaller ordinal
&gamma; &lt; &beta; such that all
r<sub>&beta;</sub>(&alpha;) are within &epsilon; of
r<sub>&beta;</sub>(&beta;) for &alpha; in
[&gamma;,&beta;). For fixed &epsilon;, this is a regressive
function on the countable limit ordinals. Thus, by Fodor's
Lemma, there is a stationary and hence unbounded set of
limit ordinals on which the function has constant value,
which we may call &gamma;<sub>&epsilon;</sub>. Since there
are only countably many &epsilon;, we may find a countable
ordinal &gamma; above all &gamma;<sub>&epsilon;</sub>. This
ordinal has the property that for all limit ordinals &beta;
above &gamma;, we have r<sub>&beta;</sub>(&alpha;) =
r<sub>&beta;</sub>(&beta;) for all &alpha; in the interval
[&gamma;,&beta;), since the values are within every
&epsilon; of each other. That is, every r<sub>&beta;</sub>
function is constant from the same fixed &gamma; up to
&beta;.

Let C<sub>&beta;</sub> be the closed interval of values s
such that the constant sequence s of length &beta; lies
within 1 of all x<sub>&eta;</sub>(&alpha;) for all &eta;
&le; &beta; and all &gamma; &le; &alpha; &le; &beta;. These
are nested and not empty, since r<sub>&beta;</sub>(&beta;)
is in C<sub>&beta;</sub>. By compactness, there is a value
s in all C<sub>&beta;</sub>. Thus, the number s is within
x<sub>&eta;</sub>(&alpha;) for all &eta; and all &alpha;
above &gamma;.

Thus, we may form the desired sequence r by using
r<sub>&gamma;</sub> up to and including stage &gamma;,
augmented with the constant value s at the stages above
&gamma; up to &omega;<sub>1</sub>. This sequence is
continuous, and it lies within 1 of every
x<sub>&eta;</sub>, as desired.

I guess this argument generalizes easily to other C(&kappa;) for ordinals &kappa; having uncountable cofinality, so that Fodor's lemma still holds.