1. The scenarios illustrated in the piece seem horrible but are biased. Consider this: a VC invests $20 million in a business and acquires 50% of it, valuing it at $40 million post-money. Assume the remaining 50% is owned by founders/key employees. Five years down the line, the company hasn't done so well and is sold for $10 million. Without a liquidation preference, the founders and key employees get $5 million: a decent payout for losing $15 million of investor money.
2. Fred Wilson repeatedly makes the point that when VCs invest in a company at a valuation of $X million, they are not valuing the outstanding stock of the company at $X million the same way a public market would. Instead, what they are doing is buying a bond + an option. The liquidation preference is the bond part of the equation: the investors are betting that the company is worth at least its liquidation preference (which, at 1X, is the amount they invested). The upside that the company might have if it succeeds is the (deeply out of the money) option.
3. Today, 1X liquidation preferences are standard and widespread, and entrepreneurs have it good. If you want your eyes to water, go read/ask about liquidation preferences in term sheets after the dot-com crash ;)