I've always loved Basic. I've always found it pretty easy to learn and intuitive to use. Used early micro versions, Sinclair Spectrum (think it was Timex in US?) and the PC versions up to Quickbasic, in which I wrote a terminal emulator that worked very well.
Did the early versions teach you bad habits? Yes. Can you teach yourself good habits? Of course. I'm still very much an amateur programmer but still use BASIC - I've been having a lot of fun with PICAXE chips recently. The fact that they are available has probably been a factor in my not having learned C yet, so there is one thing you CAN blame BASIC for :-)
I didn't write a BASIC interpreter, but I got as close as you can to that. BASIC is the reason I joined HP's fabled Calculator Products Division in Colorado. In college, I learned Algol. Couldn't stand it. Could not stand punched cards either. I could see that computers were going nowhere with batch processing. Then the university got an HP 9830 desktop calculator. You programmed it in BASIC. It had a keyboard, display, mass storage (digital cassette tape), and a thermal page printer. It was 1974 and it was clear this was the future of computing. I joined the HP division that made that machine when I graduated in 1975.
I did work on the HP 9825's HPL interpreter. Think of HPL as BASIC with the vowels sucked out. Meanwhile, HP BASIC began to rapidly advance. Labeled GOTOs, subroutines, and subprograms. Long variable names. Multidimensional arrays and matrix math. Advanced I/O. Instrument control. Interrupts handled at a high level. Graphics (monochrome then color). That all happened between 1972 and 1978.
Dartmouth BASIC was indeed limited, but BASIC had a long a useful life.