Absolutely a matter of scale. This is much less obvious than the other example - but that's because the other example was deliberately designed to be 1 in a million to make it screamingly obvious. As I said in my answer to that other question, in order to show anything with "less obvious" cases that will happen in real life, you have to show a pattern of "lucky guesses". This is the first link (only) in the pattern.
There are two basic styles of defence, active and passive. Part of the skill of the game, especially on the opening lead, is to work out which one is best for this hand. Again, as I mentioned in the other question, the best of the best make a trick-conceding mistake here one time in 5; if someone could somehow get that down to one time in 8, say, just that could vault an ordinary world-class player into a multiple world-champion.
Active defence means "pushing declarer around to set up tricks you're entitled to before she sets up hers". Passive defence means "play as safely as possible, avoid giving declarer a trick she wouldn't get without help, and force her to generate the tricks she needs on her own".
The time for active defence is when declarer might be able to set up her tricks and not have to give you yours if you don't set them up in time. The risk of active defence is giving declarer a trick she isn't entitled to. The time for passive defence is when declarer likely doesn't have enough tricks on her own, provided you don't give her any. The risk is that you may have enough tricks to set the contract, but you don't get to them in time.
There is no single right solution, but the auction gives hints (and seeing dummy gives more hints, and partner's signals help, and declarer's play pattern also helps. But by then it might be too late to switch).
Here, the fact that they have stopped low (especially in 1NT, for reasons too complicated to get into here) and the fact that your hand isn't especially good, means that partner is marked with some cards, but the opponents likely have more between them than your side does. It is likely that partner will have "honour doubleton" (Qx or Ax) in hearts, just because he'll have "some cards" (as opposed to if the contract was 6NT, needing 12 tricks instead of 7, where if the opponents are to be believed, partner will have no card higher than a ten). If he does have honour doubleton, then leading a heart will set up the rest of the suit as soon as declarer is forced to take the other honour. That's 4 of the 7 tricks your side has to take to set 1NT, and you have cards in the other suits that will get you in to take them.
So that, in itself, leans the balance in favour of the active lead. The downside, of course, is when partner has two small and the declaring side has both A and Q. Now, you're giving them two tricks where, if declarer had to lead hearts, she would only get one. Which is right? There's no way to know on this hand; you just have to play the odds.
If you're going to go passive instead, best to lead a club (the other suit the opponents haven't explicitly stated they have some of). It still has risks, but that's the least likely lead to give up a trick. It is not likely to develop any for your side, of course, you're just hoping declarer can't find a way to 7.
Now that you have your options, what do you do? Well, now we go to the quote in your question - he thinks that 99/100 would go with the heart. So does Forget... I might be a bit less secure here - there are reasons beyond the scope of this question that I'm more likely to lead passively - but I'd still guess that maybe 80/100 would choose the active lead. The chance of success is high, and the reward of success is also high; and the cost when it's wrong is acceptable (details of the Risk-Reward analysis is fascinating, but well beyond "non-bridge audience").
This time, the opening leader led passively. And this time, it turned out that the active defence would cost. That in itself doesn't say anything obvious - it's an unusual lead that was right this time. I'm sure I make tens of those a month, and even experts who are much better at opening leads than I likely make "not normal" leads 5ish% of the time - and 40% of the time the 40% play is right (and the 60% play is wrong)!
But if we can show 20 hands, say, where there were unusual actions taken, and 17 of them were "right", and the other 3 it didn't matter, but would have with a different lie of the opponents' cards, then - well, maybe that's suspicious.
Again, the defence to this is "look at all these 'unusual' leads we made that did cost us! We're just differently aggressive to 'standard expert', it's our style, and it works for us." So this is enough to set suspicion - only by analyzing every hand can we move from suspicion to "nobody gets it right that often without help".
Note: The other answers (especially Forget...'s last paragraph) explain why a spade or a diamond, even if they knew it was the right lead "somehow", would be too suspicious even to attempt.