While you might not notice any lag there will absolutely be lag introduced with that approach, it's just a matter of how much and if it's relevant to the users of the end product. Heck even the buffers are adding lag it's just on a micro-second scale, but something like an ATMega would be adding lag on a milli-second scale which is definitely gameplay affecting territory.
I can tell you from experience that if you're only a casual player of a game lag isn't noticeable unless it's really REALLY BAD. however if you're a very experienced player of a game and can play it at a high level, you know immediacy when even a few ms of lag has been introduced into the system.
I've played Street Fighter II a ton over the years, but I never really played it seriously and if you put me on a setup that had some lag, I probably wouldn't really be able to tell.
But I've put in a lot of time on UMK3 and KI2 as I own original dedicated cabs for both of those games and for a time I would play them almost every day. and on those games I'm used to being so finely tuned to the action in the game that playing those games on an emulator or even on original arcade hardware but through a "lag free" LCD setup I can feel the lag in the games reactions immediately, all the timing windows are shortened and counter actions that I can normally hit 100% of the time drops off to a coin toss.
The "I don't feel any lag" test I see posted all the time. But IMO, if you're playing on original arcade hardware it's because you care about gameplay accuracy and if you're going to talk lag, you need to setup actual tests and confirm it by numbers as opposed to just perception.
just my 2 cents.