As explained to me by someone at the Long & McQuade Music Store here (They make Yorkville and Traynor) if you run a low power head into a cab too large you will need to run the amp at too high a power for too long and either cause strain on the amp or damage the speakers because the amp is constantly sending a clipped signal to them.
Balance.
This is obviously wrong. A cab that's "too large" doesn't force you to turn up your amp. I think you're confusing power handling with sensitivity, and even then the claim wouldn't be true, because whether you have to turn up your amp too high or not really depends on your playing situation (ie, your volume needs in context).
You could use a 200 W amp into a 1000 W cab for decades without harm, if in your own context that rig gave you more than enough for what you were trying to do. For instance, if you were backing up a cabaret singer at a coffeehouse (don't laugh, those gigs are real).
Look at it this way: Suppose you have an 8 ohm cab of sensitivity X and frequency response Y-Z, rated at 1000 W. What safety do you think you would gain by substituting a theoretically equivalent 8 ohm speaker, also of sensitivity X and frequency response Y-Z, but rated at only 500 W? Would you be louder? No. Would you be safer? Also no.
Here's another thing to think about. I usually use an Acme Low B-2 2-10, which has a relatively low sensitivity. That's not a problem, because I don't play loud. Now, I used to have a B-4, the 4-10 version, which has twice the speakers and twice the power handling (700 W versus 350, IIRC). By your theory, using the bigger speaker ought to have strained the amp more. But in fact, because the bigger speaker was also more sensitive, it was louder for the same amp setting. This meant that I could set the gain lower to get my desired volume, which in turn meant that with the bigger speaker there was actually LESS strain on the amp, not more. IOW, the opposite of what you were told.