One of the problems when you discuss strategy and meta is that most people blur the lines between the two.
If you plan something based on opposition research (like at the pro esports level as an example), you are approaching that game with a specific strategy in mind. Your chosen strategy should keep the meta in mind but the strategy and the meta are distinct.
The meta comes from knowing the various strategies that are effective against other teams/players who utilize the same general pool of strategies. Each strategy in the pool comes about because it works with reasonable reliability against the other strategies in the pool. That's why unexpected plays, 'cheese,' or 'hard-counters' come into play at the strategic level (often saved for specific opponents or situations). The meta is more about knowing what situations you're likely to encounter. You know the meta, then you optimize your strategic approach for your unique situation (such as the skill-set of the various players on your team, specific tendencies of the opponent you're preparing for, or even choosing not to optimize to win your next game to keep a strategy hidden (among other reasons).
Ultimately, the strategic evolutions of iterated games is a ridiculously deep and nuanced topic, making it difficult to really get into the meat and potatoes in forums. Maybe I should make a blog post or something, but I don't know how interesting it would be.