Sometime ago, out of curiosity, I did a detailed anaysis of the hints given in early tree I had put on Ancestry. Taking a sample, covering people born between about 1700 and 1900, I checked all the hints I was offered against what I believe is correct and for which I have primary sources - a mix of certificates and parish record data. The overall statistic is that about 50% of the information in the hints was right, 50% wrong. Only about 20% of the people had a set of data that was, in my view, wholly correct and could have been imported en bloc.
Buried amongst the dross there can be useful information. I do still occasionally scan hints, and still do sometimes find useful nuggets. With a litttle practice it's not to difficult to recognise the symptoms of the "size matters" brigade, and to spot the errors that propogate from tree to tree. For some of the earlier history of some of my lines, I have a specific set of markers ( akin to DNA genetic markers !) that, when present, immediately raise the "probably rubbish" flag.
I will also admit that cross-comparing my data with hints and other on-line trees has identified cases where either I'd made the wrong connection, or my initial anaysis though correct was not sufficiently rigorous to eliminate an alternative.
Lastly, remember that all these providers of data / tree software are in the business not out of altruism but to make money. Convincing people "it's easy" is part of the marketing. And making it easy to copy others' data is part of that. Yes they do say that the onus is on the importer to check, but it's obvious many people don't.
Real research, as you are discovering, can be slow, tedious, and sometimes very frustrating. But there is the reward - and I think it particularly applies in this forum - of the personal satisfaction of a job well done.
So overall my advice would be to keep looking but treat anything you find with caution.