The Wi-Fi worked, but that did not mean it worked well.
A different router had been tested in the same place and was clearly doing a better job. That made the performance question easy. The only catch was a connected lock that I needed to reach remotely.
Changing the router could have turned one small problem into several new ones. The lock already knew the old network. Other devices probably did too. A faster connection would not be much of an improvement if the rest of the setup stopped working.
The test kept that from becoming a project. The replacement router used the same network name and password. The lock appeared to reconnect without trouble. I asked for one more remote check, and it worked.
From there, my answer was simple. If there was a specific router with firsthand proof behind it, I was happy to buy that one. If there was an even better model worth using, I wanted the recommendation. I did not need to turn a router purchase into a week of comparisons.
Somebody noticed a weak connection, tried equipment that worked better, and preserved the one part of the existing setup that mattered. I checked the lock from somewhere else and asked what to buy.
There was no reason to defend the old box just because it was already there. There was also no reason to replace everything around it. Keep the network details, keep the lock online, and use the router that handled the actual job.
My last question was the practical one: which router should I buy?