The centralization dilemma faced after scaling up is always unavoidable. Some platform products cannot keep up with user and data growth, and this issue indeed exists.

I have experience with several prediction platforms. Kalshi's market depth is quite good, with a relatively large user base, and the team's attitude towards user feedback is also quite decent. In comparison, the Poly platform's depth is also acceptable, but the team's style is more rigid, and emails often go unanswered, making it difficult to get responses to user feedback.

Although the BSC ecosystem is full of various spam activities, on the other hand, the team is more approachable, and feedback issues are generally responded to fairly quickly—though the cost of this quick response is that the market depth is truly limited. Many ideas that users want to implement on the platform ultimately give up due to liquidity or functionality restrictions.

This also reflects the dilemma faced by Web3 platforms: whether to prioritize depth or user experience, which is hard to achieve simultaneously.