sfeik sfeika l'hachmir and the bracha of hatov v'ha'meitiv

Awhile back I wrote about the issue of sfeik sfeika l'chumra with respect to counting sfiras ha'omer during bein hashemashos. If a situation arises where if A is true the result is asur but if B is true the result is mutar, the standard rule of sfeika derabbanan l'kula applies. What if we add an additional safeik to the mix? If A is true the result is asur but if B is true the outcome depends -- if C is true then the result is mutar, but if D is true the result is asur. In this case there is a 75% probability of issur, not 50-50. Nonetheless, as discussed in the previous post, R' Akiva Eiger says one can be lenient even in this situation.

I happened to see an Aruch haShulchan that left we wondering if there is more to this issue. If a second bottle of wine is served at a meal a bracha of hatov v'ha'meitiv must be recited (see O.C. 175 for the relevant details). There is a three-way dispute in the Rishonim as to what type of wine this bracha applies to:

1) Rabeinu Tam -- any wine, as the bracha is on the new bottle, not the quality
2) Rambam, BaHa"G, Rashbam -- only on better quality wine or a different variety of wine
3) Rosh, Tosfos -- any wine not of poorer quality than the first

The Shulchan Aruch (175:2) rules that as long as the second bottle is clearly not of inferior quality than the first, even if there is a doubt whether it is in fact better, a bracha may be recited.

Why did the Shulchan Aruch not rule like Rabeinu Tam? It would seem that this accords with the well known rule of safeik brachos l'hakeil -- when in doubt, no new bracha is recited. Since Rabeinu Tam's view is in the extreme minority, the chance that he is right is not sufficient to warrant a new bracha. However, what is more problematic is understanding why the Shulchan Aruch rules like the Rosh and Tosfos and not like the Rambam. Based on the principle of safeik brachos l'hakeil we should not recite a new bracha unless the second bottle is clearly of better quality so as to satisfy even the Rambam's view. Why chance relying on the Rosh and Tosfos when it may lead to an unnecessary bracha?

The Aruch haShulchan answers (175:4) that this is an example of a sfeik sfeika l'hachmir. There is a 50-50 chance whether the second bottle is in fact better. Even if the second bottle is not better, in fact, even if it is worse, one still has Rabeinu Tam's view to rely on that any second bottle requires a bracha. This second safeik tilts the odds in favor of reciting a new bracha.

Is there a possiblr difference between safeik brachos l'hakeil where according to the Aruch haShulchan the rule sfeik sfeika l'chumra applies and sfeika derabbanan l'kula where it doesn't?