For instance the word "Daughter" which is the same in most Germanic
languages, shows a completely similar pronunciation between Swedish
and Persian,( dotter= dokhtar ) as if they have learned it from each
other.
Two things cannot be "completely similar": they can be "absolutely identical", and they can be "similar". You can apply various scalar adverbs to "similar" such as "highly", "fairly" or "somewhat". The words dotter and dokhtar are similar, also German tochter is similar ("equally" similar is being more similar in one respect and less similar in another). This is not because modern Swedes (or Germans) and Persians get together for word-swapping, it is because they learn their respective languages the normal way people learn languages, with words being passed down from generation to generation, and this word has not changed much. The original source word was indeed a single word, used thousands of years ago. This is not a coincidence, it is a well-known and well-studied pattern. If you listen to a speaker of Swedish and Persian saying the respective words, it is obvious that the pronunciations are not the same.
Whether or not any other words on your list are similar by coincidence vs. similar because of common origin can only be determined if we have a more objective and informative data source. Swedish bror means "brother" (I don't know is there is a different word that you have in mind). Not knowing Persian, I can't guess what "berar" is except by association with bror, and I can at least determine that there is a Farsi word برادر conventionally transliterated as berâdar or berâdar which likewise descends from a common source in proto Indo-European.
گوشت [goʃt] and kött [çøtt] are only somewhat similar, but fortunately we know from historical study of Swedish that originally the consonant spelled k was also pronounced [k], which has an unknown historical source. The Persian word seems to be similarly of unknown origin. It is possible to conjecture that proto-Germanic *ketwą or the verb *kutjaną "cut" is related to this Persian word (the correspondence between Germanic [k] and Iranian [g] is known). It could also be a coincidence.
The null hypothesis is that any two words are historically independent, but you can set aside the null hypothesis if you have sufficient evidence for an alternative. There are numerous criteria used to support a hypothesis of historical relatedness. First, there has to be a consistent chain of connections in form going back to the putative proto form. Directly comparing Swedish [brʊr] and Persian berâdar runs into a problem regarding the d, but fortunately we know that some dozen centuries ago is was pronounced bróðir in Scandinavian. Second, there has to be a credible semantic chain between the words, meaning that known earlier meanings of words have to be taken into consideration, and you cannot just look at modern Farsi, you have to look at all of the related words in older languages (Old Persian, Avestan, Sogdian) and sister languages (Luri, Kurdish, Pashto). The connection of mellan and miyan via PIE *médʰyos is also not beyond the pale even if it is not a conventionally-accepted truth.
The connection between هاپر halpar and Swedish hoppar – I guess "hopper" – requires lots more evidence in order to overturn the null hypothesis of unrelatedness. For example, is هاپر a bin into which you would load grain for grinding? In other words, there is always the conceptual possibility of two words being historically related, but the burden of proof is on you to establish the chain of evidence that overturns the null hypothesis. Direct comparison of isolated words in two modern languages is of negligible value in providing that proof.
