The two questions you mention are indeed both poorly suited for the Stack Exchange platform. The problems they have are rather different, though, and some of them may be fixable.
The first question was closed as "too broad", but IMO, a more accurate closing reason would've been "unclear what you're asking". Basically, condensing the question and its answers and comments into a few lines, they essentially went like this:
Q: "I want to make a running game with infinite ground. How could I do that?"
A: "One way is to split the world into several parts, and keep creating new parts in front of the player and removing old ones behind them."
C: "I tried that, but the player falls through the floor! Why does that happen?"
Basically, it's a kind of a chameleon question — it started out as a conceptual question about maintaining the illusion of an infinite world, and then morphed halfway through into a "there's a bug in my code" question (with no actual code included, which makes finding the bug extra difficult).
Questions like these cannot really be meaningfully answered, since the actual question being asked is a broad, nebulous and moving target. To try to give a useful answer, one would have to guess which level of abstraction, from general game mechanics to specific coding syntax, the OP is actually having problems with — and it may well turn out that the answer is "all of them", in which case the question is indeed too broad for Stack Exchange.
The general way to fix such questions, where possible, is to engage in comment dialogue with the asker to try and clarify what they already know and what they're specifically having trouble with, and then edit the question to focus on that one, specific, answerable issue.
In some cases, it may also be possible to split such questions into multiple parts, one for each of the component problems. For example, this example question is broad enough that it could be split into (at least) three distinct questions with distinct levels of abstraction:
"How can I create the illusion of infinite ground in a running game?" (abstract, platform-independent),
"How to implement infinite ground in AndEngine?" (concrete, platform-specific), and
"Why does the player fall through my infinite terrain?" (very specific; should include a minimal example for testing).
As for your second example question, the main problem is simply that it's phrased as asking for the "best" way to do something, even though, underneath, it's really a "how to" question.
To see the difference, compare these two hypothetical cooking questions:
Bad question: "What's the best way to cook an egg?"
This will invite a countless list of subjective answers, with one answered preferring their eggs hard-boiled, while another one likes theirs fried sunny-side up. (Cue lengthy comment debate on the relative merits of sunny-side up vs. over easy.)
Good question: "How do I cook an egg? I tried putting it in the microwave, but it exploded!"
Good question: "How do I cook an egg? I know I can just boil it for ten minutes, but I never get a nice runny yolk that way."
These questions clearly specify what the asker wants, and how far they've got on their own (i.e. in development terms, which level of abstraction they're interested in / having problems with). While the topic is still very broad, it's clear that the specific question can be adequately answered in a reasonable space — at a bare minimum, "immerse the eggs in boiling water for three minutes" is technically an adequate answer.
The one thing these good questions don't do is ask for the single "best" way to do something. Somewhat paradoxically, by not doing so, they actually encourage comprehensive answers that describe several good solutions and compare their respective advantages — like, say:
"There are many ways to cook eggs, but, as you've noticed, microwaving is generally not one of them (although it can be done). One of the simplest ways is to immerse the eggs in boiling water for a short time; depending on the immersion time, this can produce either a fully congealed ("hard-boiled") or a partially runny egg. [...] Eggs can also be fried on a hot pan. Again, there are several ways to do this: [...] Other methods of preparing eggs include shirring, poaching, scrambling, etc. [...] Eggs can also be used as an ingredient in many dishes, such as..."
Addendum: Prompted by your comments, I tried to see if I could edit your question into something that I would personally find useful and answerable. Here's what I ended up with.
However, while doing so, I ended up repeatedly running against the fundamental issue that your question doesn't seem to include "a practical, answerable question based on an actual problem that you face". Specifically, based on the impression I got while editing your question, I strongly suspect that you asked it before even trying to just make even the simplest 3D model (say, a yellow sphere) in any modeling program (say, Blender) and importing it into any game engine (say, Unity).
At this point, my main advice to you regarding this question is:
If you have tried that, please describe what you did (briefly — no need to go into details) in your question, so that potential answerers have some idea of what you're familiar with and what needs to be explained.
If you haven't tried that, go do it now! Then come back and, if you still feel that there must be a better way, clarify your question based on what you found easy and what you found difficult.