Kinda? You can find videos of any car catching on fire, what matters is how they compare at a statistical level with their competition to see if they are doing better or worse, not evidence that at minimum failure rates are nonzero. I truly don't trust China Watchers to produce anything meaningful or useful.

Define "plenty." What's the fire rate compared to the total number of BYD cars in existance? How does that compare to Tesla? How's the historical trend, is it going up or down? Are they equally likely to catch fire across the board or are there problematic models or series? Different countries have different automotive regulations and therefore have slightly different cars even if they're the same model, are the fire rates different by country? Are domestic Chinese BYDs more likely to catch fire than cars exported to, say, other Asian countries, Latin America, or Africa?

This is why you don't draw conclusions from how many YouTube videos you can find. When there are billions of any product there will inevitably be tons of videos of it going wrong which doesn't inherently tell you whether it's actually likely to happen or not. If that was an acceptable statistical analysis method, then I can binge watch the hundreds of thousands of aviation accident videos and conclude that literally all planes do is crash.