The most important improvement in T18 is the thickness of the base of the tip. If you put calipers on it, the 900M tip is 1/4" in diameter. Exactly 0.250". The T18 tips are 0.267" in diameter. This allows for more copper filling for better thermal coupling between heater/sensor and the working tip of the iron.
My eyeball calipers say your generic tips are roughly as thick as T18; the difference would be very obvious if the clone were only 1/4". The original form of the 900M also have an obvious taper/bevel around the base which your tips are lacking. This looks like a clone tip that has deviated from either spec. Maybe as an improvement, having the thickness of a T18 tip but retaining the length of a 900M tip? Certainly, I can't see that making the tip barrel/cavity a smidge longer would decrease performance, here, but I don't know what the inside of the Hakko heater looks like... I suppose the extra 1 or 2mm could end up dissipating heat rather than collecting it. The hole being larger/looser though? That is a drawback - possibly a compromise so that it works with all clone stations (reduces customer support costs!). But another factor is cost for making new dies/molds. Even if you're an expert in all the processes, it is very difficult to predict the final dimensions down to micrometers. And making new dies and samples can cost tens of thousands. In general, the clones will usually err in being slightly loose for completely practical cost reasons. If it's even slightly too tight, it's garbage. I only have 1 genuine 900M tip (I hope it's genuine; $20.00 shipped from Japan, lol), and it's only slightly looser than the T18 tips. I have a couple Plato brand 900M tips, and they are noticeably looser, yet. ALL of these 900M tips, genuine included, have noticeably lower thermal performance than genuine T18 tips.
But the real story is how much copper did they stuff in there? Does their process leave air pockets/gaps between the copper and iron layers. Is the iron layer consistent in thickness. Is the iron alloy wettable and resistant to oxidation as the real deal. Is the chrome plating as durable. Is the tip solidly constructed and brazed to prevent separation and oxidation of the copper inside through thousands of heat cycles?
If your clone tips aren't performing well, it could be as much to do with what you can't see as what you can actually inspect and measure. The more accurately you can cast and mill the iron jacket (possibly adding time/cost), the more copper you can stuff in there without leading to structural failures. The iron alloy is very important, too. It's a compromise between structural strength and stiffness*, thermal conductivity, wettability, corrosion resistance at high temp, how it casts or MIM's, machinability, and how well chrome can be plated to it. Hakko tips are well regarded for wetting like they just can't stop and for lasting a long time.
*I read of a guy dropping his [other big name] iron, and the tip bent 45 degrees. I've dropped my 1.5mm bevel tip which point down, and I just curled the lip of the tip on the very point. I pushed it back and still using it. It is a TFO, and the chrome didn't even crack or flake beyond a tiny spot lost right at the deformation. A cheaper tip (like my clone T12 tips) would have a huge crack and area of delamination/flaking.
In short, there are a lot of compromises in a tip. And a lot of processes involved. It is easy to copy a tip, in appearance, without spending all the time that Hakko has to fine-tune and optimize every operation and even the iron alloy, itself, with eye on the end result of the sum. And it's not necessarily that Hakko decided to spend a fortune in R&D. They have been making irons so long they have the benefit of many natural iiterations of replacing tooling to fine tweak the formulas, materials sourcing, and die and machining specifications/diameters and all, along with customer feedback and complaints. The first genuine Hakko tips were probably worse than the modern clones.