No, there is no danger of restarting postgres using the init.d method.
However, to restart it because something weird happened is not a good idea, because it limits severely the amount of information you can collect to find the root cause and limits opportunities to fix it.
Also in all the years I have been involved with postgresql, I never encountered a situation in which a restart “fixed” the problem. The immediate ‘incident’ might be resolved but if there’s a problem, it will still be there.