`curl -o /dev/null -m 222 \"$Link\" 2>> $logFile`;
    my $curlResult = `echo $?`;
[download]

That code is quite convoluted.

Each `...` call starts a separate shell, so the second shell doesn't know the return value of a program started in the first shell. And even if it would, you'd not ask it for the return value of the previously launched program, because you interpolate the perl variable $? -- which you could use directly anyway.

But there are more robust ways to interface curl anyway, for example WWW::Curl and LWP::Curl.