Anyone who thinks RPGs, regardless of format, would cause the TG-16 to be massively more popular evidently doesn't remember that most console RPGs at that time sold in very low numbers. Square skipped FF II-V, and every other RPG that had a Japanese release during that time. Enix skipped Dragon Quest V and, IIRC, left the US completely after Ogre Battle came out.
It really wasn't until FFVII, long after the Turbo was dead, that RPGs became mainstream enough in the US to effect the success of a system. In 1990 RPGs were basically a money loser. Think of the Genesis, which made money hand over fist. How many RPGs did it have? Not many more than the Turbo. Maybe less? The SNES, equally successful, had a couple more. I think even if you used the looser definitions of RPG to include Shining Force, the 1st person dirges, and even action RPGs like Zelda and Oasis. What are we looking at, maybe 50 RPGs during the entire lifespan of SNES and Genesis combined? And maybe 4 or 5 of them are best sellers?
Never mind this top ten list (which was compiled...how, by who?). If anything I'd say the shooters helped the system since all my friends who moved on from 16-bit ages ago probably have a better memory of Lords of Thunder than any other game.
What killed the Turbo, in addition to the high price, was that the sorts of games that eventually came to define it in Japan, the stuff it really kicked ass at, the sims, the RPGs, the anime licenses, the digital comics, this stuff...it wasn't something you could bring to the US. You'd spend 100x as much money localizing a game 1/100 as many people would want to buy. They games that did sell in the US (Madden, Mortal Kombat, Mario, Sonic) don't have very good representation on the PC Engine. Sure, the system loyalists here seem to think Blue Blink is as good as Sonic 2, but most people don't see it that way (US or Japan) and marketing has nothing to do with it.
It just wasn't meant to be. The only thing that would have saved the TG-16 (aside from Americans just spontaneously deciding they'd like to read in a video game) would have been if NEC/TTI suddenly started producing dozens of grade A original software for the US, which I'm not sure they could have done with any sum of money.