I don't think that's the case. 2/3rds fail, 2/9ths return a small multiple, 1/9th return a substantial multiple to a typical decent VC. Valuing the shares at half the most recently priced round puts them into a quite reasonable range, IMO.

The company's valuation at the last financing round already prices in the high likelihood of failure. That is, on average, the VC industry does a good job at assigning valuations to companies (see published reports on average IRR of the VC industry). As you point out, there's huge variance in outcome which may affect your appetite for shares, but when it comes to assigning an expected value to your shares, the last financing round is an appropriate place to start.

I don't know what your experience is, but what I have observed is that valuation is usually the product of multiplying how much money investors want to put in by how much stock the company is willing to give up.