Aside from the technical issues, for a significant increase in players per server there would need to be a change in combat mechanics to take variable net latency into account, and there would need to be vastly increased anti-griefer mechanisms added (which adds an entire worms nest of problems). Similarly for any in-game economy to arise beyond the current adhoc trading between friends would, especially for larger public servers, require restrictions to be available for what gets brought in to a world - including characters that have been enhanced artificially. It would be a big mess and quite frankly, without those factors all taken into account from the very start, a programming nightmare.

As for balance, they seem to be addressing that as time passes. I'd personally like to see more alternatives in progression, land transportation, and, above all, I think everyone would like to see more content.

One interesting possibility that occurs to me is incorporating some variant of the old roguelike bone files - some automated process for creating content by transmuting aspects of player world changes, from volunteer players of course. It would be neat to stumble across the ruins of what was once someone's base - perhaps infested with mobs from the biome and containing some random drops in chests that were once player made content. Could even have the 'made by' tag still attached for fun.

But beyond a vanilla, please everyone (or as many as possible) balance measures, I expect quite a bit of game changes to continue to be available from third-party mods.

So no, MMORPG is right out, and any significant increase in players per server would soak up a lot of developer and programmer time that is needed elsewhere to bring us more of the developer's vision.