The difference is that some people, like me, actually put forth a number for things like the chances that Treasury converts (75%) or writes down (25%) the SPS. Most others just list possibilities without assigning numbers to them, which is completely useless when deciding whether to buy, sell, or hold at current prices.
That's just the endowment effect, a well-documented phenomenon.
Even I would sell some juniors to buy commons if the prices diverged enough. The FNMAS:FNMA ratio hit 4.0 today; if it goes much higher I probably will do just that.
The 6.8% figure depends pretty heavily on if you're buying the most liquid series (FNMAS, FMCKJ), the least liquid (usually the 5-letter ones starting with FMCC), or somewhere in between. They have all done well recently though.
You can only position yourself one way. It will be a function of the different possibilities and probabilities you assign to those variations.
I suppose that depends on what you define "winning" to be. If the juniors outperform the common you would lose relative to owning only juniors and no commons. But differences of opinion are what make a market.
I don't think either class gets totally wiped, but a SPS conversion would cause there to be little upside, and perhaps substantial downside, from current prices.