we are talking about period when indo-european languages spoken in europe werent as distant from each other as they are today. i dont know about any convincing evidence that would allow us to clearly separate language groups of peoples known as illyrians, dacians, etc. consider the closeness of italic and celtic languages.
by the way, toponymy of beskids may come from old german besche "roztok, watershed".
then there is question of origins of hill names like vihorlat, hirlata, hoverla (the last probably comes from old hungarian, idk whether it is confirmed), and i think horhany falls into this category as well.
then we have hydronyms such as hornad (older chornad, in śpiš german – kunnert), udava & ondava (funny nasal game there?), torisa (hung. tarca), but then latorica & laborec (hung. latorca, laborc)…
but then again, this would be heavily off topic :)

South Slavic grad, Vlah
East Slavic gorod/horod, Voloh/Valah
West Slavic grod/hrod, Vloh (Vlochy = Italy in Polish)

polish gród (grud) (town), czechoslovak hrad (castle)
valach is either latin or hungarian influence (or both) afaik; we say valach, but walnut is vlach nut here.

West Slavic *melko > mleko/mloko etc.

czech mléko, slovak mlieko, mlíko, slovjak mliko (but here in southwest we say mlěko… i mean, southwestern part of the eastern slovakia)

You are correct. I forgot to mention that Czecho-Slovak behaves dually in this feature.
For example, they say grad/hrad, but robota (hence Robot) instead of rabota and rozumet rather than razumeti:

yes, slovak robiť and slovjak robic -> porábať, porabjac, however, slovak kráčať and slovjak kračec -> prekračovať in slovak but prekročovac in slovjak.
also, slovak chrbát, but slovjak chribet.

consistency? we dont know such a pokemon :D