I'll take a stab at answering some of your questions,
Nestle;
Yes, I found fellowship to be more demanding than residency... I'm at the hospital more, fewer days off, and no leaving early post-call. However, it is also more rewarding. Also, call is better (home) and when I do come in, it's for something interesting (eg: tranvenous pacer, pericardiocentesis, echo, ST-MI, etc...). In our program, we cover two hospitals and take q6 weekday call and q3 weekend in 1st and 2nd year, then 1 weekday call/mos in 3rd year.
Most of our clinical resposibilities are in the first year of fellowship, prob'ly harder than internship, mostly because you are the 'go-to-guy(gal)'. Everyone on the wards or units needs cardiology for something. But it's great being able to make the diagnosis, do the procedures and offer the definitive treatment that the other services want...
Echoburress;
Of course, the amount of research that's 'good enough' at one program may not be at another. A simple rule of thumb would be to do something and try to get published. If you are making a concerted effort to contribute to research in some fashion that'll help. Even if you are not published, list this on your CV under a heading titled "current research": 1) the title of your project, 2) under who's direction, 3) at what facility, and 4) dates of involvement.
My best advice regarding research would be to get started early, you can't start soon enough. In fact, there are 2 (TWO!) 4th year med students who come to our interventional research meetings each week. Plus, they both had abstracts at the AHA this year and have submitted manuscripts (one' first author on a manuscript in Circ already). Of course, these are the guys we all hated in med school, making the rest of us look like slackers. But, they are going to Bringham and Womens and Mayo next year and will be able to choose what Cards program they want...
So, even if you didn't start research in med school, start now! Ask a cardiology fellow or staff about their research and see if you can help. You won't be first author for the first paper, but often you can spear-head a sub-analysis and take first author on that. I suggest doing retrospective research. Prospective will take too long for it to produce publications for you. Also, you could come up with your own idea for research, but that's not easy... Best is to help develop a database (eg: I put together a procedural database and have been publishing from that). You'll find that once you get started with something, you'll be able to do multiple anaylses with that. It's not hard to put together an abstract, submit it to a nat'l meeting and write up a manuscript. You'll want to work with someone who's published and knows how to write and can proof for you... Good luck.
Mollie;
I didn't do a rotating internship. I went straight to ACGME IM residency. If you want to work in one of the 5 states requiring DO internship, you may want to do this. Although, I've heard, there are ways around that rule as well. If you do do an internship, I'd do it as part of a track for IM residency, so you don't loose a year. Most ACGME programs won't give you credit for the internship if you leave the AOA after your first year though (but, you could do internship as part of a dually-acredited AOA/ACGME program).
I'm not sure if doing a DO internship then an ACGME IM residency would make you eligible for DO cards fellowship. You would be eligible for ACGME cards fellowship though... (Again, if you went to a dually-acreditted program, you should be eligible for both).
That would open more options for you, however, I'd do ACGME fellowship first and use AOA fellowships as a back-up. Interestingly, as I've been getting calls from recruiters and as a DO, I am asked specifically if my training was through the ACGME or AOA. Recruiters tell me many groups specifically state they are interested in only ACGME trainees.