To the point of the original question - I think it's perfectly acceptable to determine a reasonable selling price from the Tracker data. For no other reason than the Tracker showing you actual realized selling prices for that item. These aren't collectors estimations, they are not a collector talking his book to skew the prices in a way they want. It's pure factual market data.
But the Tracker is not magic, it's not going to assign you a price that *you must sell at*, and you still have a responsibility to make a decision on how you are going to interpret the data in front of you. For example, take the below screenshot. If you were trying to determine a fair selling price for a loose complete Barada you see an average of £58. But if you look at the scatter, you can see various sold prices anywhere from £43 to £74 - as Grant said "book ends".
So what's the right thing to do? That's completely up to you now. The price ranges could indicate poor vs very good condition (but also be aware of licensee variation, COO markings etc), so you could factor that in with your item in hand. To help gauge this, each data point is "clickable through to source" so you can see exactly what the condition is for each of these sold prices. If you don't agree with a particular data point you can even exclude it right there and the analysis will update accordingly.
If I wanted to sell quickly, I'd price around £45 expecting it to be snapped up. Otherwise, I'd make a judgement on how good condition my item was and then set a price anywhere between the average and top price. Then you sit and wait and let the market react. I feel if you follow this basic recipe then you can absolutely defend any selling price if it's ever challenged.
A price guide attempting to provide a value for every conceivable condition of an item is an impossible task. But I feel the Tracker avoids this issue entirely by presenting each individual price point which can be followed to source. The rest is up to you.