LG is the only one that could be said to have sufficient marketshare to justify that investment for any but the largest VOD providers now. The revenue generated from those platforms once rights holders are paid is a fraction of what it costs to maintain them simply through testing overhead. Android TV and Tizen's marketshare has increased rapidly over the last few (read three-ish) years, and when you often need to account for international rollout or repurposing in your cost analysis it's even stronger (Roku has zero marketshare outside of the US, many large cable companies outside of the US have Android TV based set-top boxes with multiple million installations).
This generally also applies to browser based streaming, with the added bonus that is where a bunch of piracy comes from.
I'm struggling to understand the claim made by this statement. Maybe I missed something? What does HTML have to do with serving video over the wire? Underperforming on which metrics? What's the time value of "these days"?