On the valuation. Optimizely's pre-exit valuation rose in every VC round since changing the focus of the business from SMB (Series B) to Mid-Market (Series C) to Enterprise (Series D). The valuation of the exit took a hit for reasons that were not the enterprise focus. I could say more here, but consciously choose not to.
The self-serve model at Optimizely was certainly flawed. If VWO was able to build it profitably, that's likely because their labor costs, being based in India, are substantially lower than Optimizely's. Keeping a legion of engineers funded in SF has very different unit economics than keeping engineers funded in India.
Now on your last argument, that self-serve customers could acquire enterprise accounts, I certainly agree. If that strategy were better executed, it could have created a valuable pipeline. But the self-serve to enterprise funnel was sorely lacking, and when self-serve was cancelled, the business benefited financially and migrated everyone to the new pricing model for substantial success.
Its easy to look at Optimizely's final fortunes and link that to the enterprise approach, and I won't argue they're completely unrelated, but there were very different factors that caused Optimizely to fold into EPi's acquisition spree.