But anyway,Ossetians are the direct descendants of Scythians in their purest form surviving today. I suggest you read about them.
You really have a weird fixation on historical/cultural/genetic pureness.
Ossetians might be the people with more cultural ties with Alans, a sub-group of Sarmatic peoples themselves more or less related to Scythians. Except in nationalist and protochronist historiography, I never saw argued they were "pure Scythians". (If you really had to find the "last" Scythians in the sense Herodotus gave, you'd find them in Crimea where they were last identified as distinct people from Sarmatian groups)
Now, what "Scythians" meant for ancient peoples in the broadest sense : let's first remember that neither Greeks or Romans really attempted to be scientific ethnographers and had only a faint idea of the relations between peoples in European hinterland. It's why they ended calling Celts (originally a people, or more probably a regional coalition) undifferentiated groups out of remotness and dropling of records. The same happened with Illyrians (we know were a group largely defined by Greeks and their interaction with locals, that gathered an awful lot of Celtic, Gaetic, Italic or other Balkanic groups), later with Germanics and without real doubt with Scythians. Ethnographically, it had approximatly as much value as "Hyperboreans", or more surprisingly the old Thierry's hypothesis about Kimris (which, even if you're not knowing it giving it's a fairly old and local thesis, you're basically using with some genetic considerations) : after all if we define broadly enough Scythians as any IE people with more or less ties with proto-northern Iranic archeological cultures, we could as well include Cimmerians into the mix and make them the responsible of phenotype closeness (assuming that standardized description of Scythians are reliable, see below) in Europe. Of course, the problem is that it's neither archeologically or historically attested (especially as we know that proto-Germanic peoples might had a really important non-IE substrate culturally).
Now it's possible Greeks borrowed a lot of considerations on Scythians from a mix of their own interaction in Black Sea, but as well with interactions Persians and Medians had with Sakas, which was as well more of a geographical terming applied to a range of various if related peoples rather than any real description of unified phenotype/genetics or even firm cultural continuity. Think of "Scythians" as it was about "Ligures", "Barbarians", "Berbers", etc. a broad term based on experience neighboring peoples, the only that wrote them down and systematized them. It created a systematized depiction of Scythians from the Vth century BCE to the Vth century AD among Greeks and Romans whom accuracy is suspicious by its systematisation alone when we know you had whole migrations, mixing and desertions of peoples in the length of a whole millennia.
Now as for central Asian Scythian they looked like Asronan and Arteshtaran of ancient Iran or the old Brahmins of India who were known to keep a pure lineage both on the maternal and paternal side for a very long time and it's only after some time that these people began to mix with the rest of the population so the answer to the question did the Scythian look like northern European stock the answer is a resounding no, the look like old Indo Iranian people, and I don't think a lot of them were blonde may be brown some ginger and blonde hair but it's hard to know since Indo Aryans or Iranians cremated their dead or perform sky burial
The problem arise from systematized description of Barbarians of the steppe by Greeks, who not only generally assumed whoever they encountered were representative of the whole of "borderless" regions where far Barbarians lived (every people North of Celts are Celts and look alike, every people North of Scythians are Scythians and look alike, etc.) but furthermore were re-used continuously for a millenia as a proto-ethnographic trope even when very obviously you had a lot of various human groups mixing and movement.
So while it's perfectly reasonable to think peoples they encountered in Pontic coast and near hinterland had a "Nordic" phenotype (described as grey eyes and clear hairs which might mean as well blonde/reddish or even light brown) the trap there is to assume that you had a direct and unaltered phenotype/cultural or even genetic continuity if not stasis.
That said, I find equally risky and dubious at best to apply to ancient and protohistorical peoples the same practices you can see in social groups within related/later peoples : it looks convincing but implies a definite, concious and continuous ethnicisation of social classes that are rarely (at best) found historically or archeologically. Generally, in the manner of generation, such identity is the result of a quick mixing-up which is common enough that it ends (amusingly) with a large stress on a percieved and mythified ethnic origin.