Koichi Tohei got a lot of his training methods from the Tempukai, and invented the rest as a result of being challenged in Hawaii, as Yoshimitsu Yamada recounts here
. He didn't "get them" from ShinShin Toitsu, though - they were
The consensus seems to be that he took rather little of his training methods from O-Sensei.
Folks, I'm not an aikidoka (outside of IP/IS training), but have an interest in what's common to Daito-ryu and the arts descended from it more than I care about the differences between them.
Mitori-geiko: the great historical differentiator in traditional budo teaching. Clearly all of Sokaku Takeda's best students were proficient in it, as were the best of Ueshiba's, in particular Tohei. All of these standouts operated within a similar technical syllabus and gleaned and/or developed solo training methods compatible with it. This is the bigger picture. Daito-ryu is in many ways unlike preceding/contemporary bujutsu/budo. In turn, aikido, as a whole, is a bit different in flavor from, though clearly technically indebted to the various Daito-ryu(ha), as a whole. Neither Takeda nor Ueshiba insisted on their students following a rigid syllabus. Evolution and personal interpretation that still hold true to the central principles of the system, i.e. those principles that support these arts' eponymous skill -- aiki -- were clearly lauded by both men: e.g. the regard in which Ueshiba held Tohei's understanding and interpretation of aikido.
There are precious few people alive who've had the honor and opportunity to train with both Ueshiba and Tohei. My understanding is that at least two still reside in Hawaii, and one has already weighed in, saying that what Dan Harden is doing is in line with the aiki of Ueshiba. While I don't know this gentleman's specific take regarding what Dan is teaching relative to what Tohei expressed regarding aiki, if Ueshiba felt Tohei was on track regarding aiki, and people of the Tohei lineage here feel that IP/IS methodology (Dan's per se) is compatible with what Tohei taught, then all these data points support a clear relationship between the various methodologies, allowing for differences in flavor that would've been acceptable to Takeda and Ueshiba.