You can talk theories, and programming skill, and single screen tech demos made 10 years after the death of the system, but the shit never actually happened on the PCE during its life which leads me to believe that it was either impossible, or such a bitch that it wasn't worth it because there is no shortage of good programing in the PCE library.
You make a good point here, and it's one that I agree with. SD3 was never ported to the PCE, and no one ever really asked for SD3 to come out on the PCE anyway, because there were already tons of other great programs. All the theories about "how" to do it (or not do it) don't matter, because all the talk in the world can't change the past.
Although I think that if systems weren't retired to make way for new ones, that development for those systems would continue to progress to new levels we'd have never believed(look at euro & brazil SMS, the short life span of the Arcade Card, or even where the Neo Geo got to)...
-it's fine and well to say that with most major consoles, they were more or less fully utilized during their time. But the PC Engine is one of the exceptions where, maybe cuz its a bitch to program for or because of the HuCard/CD formats...
-it wasn't really pushed to very much of it's potential very often. There's more of a wide range in technical displays on PCE than on MD or SFC. Or in other words, for every Drac X and Lords', there are more technically simple (not necessarily bad)games than the other guys.
After the CD format took off, most games were made with quantity over intense quality. PCE fans didn't stop buying games that lacked parallax or 'Mode 7' when they had CD music, voice and cinemas in it's place. It sort of became it's own market like the N64 & amecube.
The Genesis however, was in direct competition with the SNES, so even though it lacked the nifty built-in effects that the SNES had, anyone publishing a top tier game for it had to really push the system to make some visually intensive games. There wasn't any reason to do this on PCE(other than pride).
More important than being able to push hardware though, is managing artwork/graphics. Square may have been great at getting nifty effects out of the SFC, but their games looked best because of how they used art. The main difference between Final Fantasy IV and SD3 or Dragon Quest V and VI is how they made the sprites and tiles. The best looking Genesis games use smart use of color to look like they're displaying double the system's limit.
If Square had the incentive to make a few killer PCE games, but wound up putting out a bunch of duds, then it would be a clearer sign that it was a matter of impossibility. But look at what Falcom and Quest did with the single PCE games they developed.
The PC Engine CD would be perfect for a later Square style game. Even if they used all static bg's and only nicely animated effects. As long as the quality artwork was there. Afterall, most of the crazy effects used for spells and such in SFC carts was a way to save on memory.
When Secret Of Mana was brought up as a dream 'lost game', I wasn't hoping for a straight port anyway. I think a PCE specific one would look at least as good overall. But if the SOM/SD games really are great games and not just eye candy, then keeping the gameplay intact should be the only concern with a port turning out, right?
But on a technical level, when someone says, "
if it could be done it would've been", I say look at what a lowly homebrewer like Chris Covell has managed on his own in such a short period of time. And not just his PCE demos.
