Eric Cantor, the Republican leader, told me yesterday that he assumed they would assume the majority in November. ~Mark Shields 
If there is one thing from the last week that should deeply discourage Republicans, it is the realization that for all of the real successes in New Jersey, Virginia and Massachusetts, their party in Congress is still led by the same people who presided over two very large consecutive electoral defeats, most of those leaders were complicit in the bailouts their constituents hate, and these leaders continue to have no correct understanding of why they were voted out of the majority. That doesn’t mean that voters know or care about Boehner, Cantor, McConnell and Kyl themselves. Voters never knew or cared about Pelosi and Reid, either, and campaigns that tried to drag down effective Democratic candidates by attaching them to their leadership never worked. That said, party leaders in Congress are not irrelevant when the party is in the minority.
These are the people who are the functioning political leadership of the GOP, and they have been unusually unsuccessful in repairing the image of their party, crafting anything resembling a coherent opposition agenda and providing the public with any reason to believe that they would handle another turn in the majority with any competence. On top of it, if Cantor is any indication, they seem to be no better at analyzing the national political landscape. If Cantor actually assumes that the GOP will win back the House in November, he is engaged in wishful thinking or has simply spent too much time listening to flattering, unrealistic claims made by other Republicans.
Republicans would need to gain 40 seats to win the majority again, and that will give them the bare minimum of 218. At most, they might realistically net 13-16 , and that is assuming that things continue to go their way. Even if every seat listed as “lean Democratic” by CQPolitics today fell to a Republican candidate, and the GOP won all other vulnerable Democratic seats, that would be only a net of 37. That would be a significant and remarkable gain and larger than the Democratic pick-up in 2006, but that is as high as the GOP wave can possibly crest. In the last
36 34 years, Presidents with approval ratings above 40%  do not normally lose 30+ seats in the House. 1994 is the one exception, and that result was greatly aided by the huge number of retirements of members of the majority in that year. At the moment, Republican House retirements still outnumber Democratic retirements.
If Cantor automatically assumes that the GOP will do better than this in one cycle, he is dreaming and complacent. That tells me that the GOP simply expects victory to happen, which makes it more likely that they are going to be unprepared to fight the election effectively and they will end up being badly disappointed.