The unknown (at this point) is the tower. The inverted-L will induce current into the tower and the input impedance will depend on the the tower electrical length. The tower should be connected to the radials.
Correct, and the other unknown (less of an impact, but still noteworthy, especially if large) is what antenna(s) are at the top of the tower. For instance, how far above the horizontal portion of the inverted L is any antenna and what antenna is it? I've found that this does matter and it should be factored into your models.
Whatever length of wire you decide to use for your new Inverted L, it might be quite helpful and instructive if you could have a vector analyzer present to assist in the design of your matching network. Every situation is different and even though yours and my inverted L may be very similar in physical characteristics (height and length of horizontal section) what we have in the way of towers, antennas on the tower, and objects adjacent to the near field play an important part in its design and final performance. Mutual coupling is not unlike mutual attraction; one needs to be careful
GL, 73, Rich, K3VAT