Yes, I1 today, descend from one male survivor who lived some time around 2600BCE. There is of course older I1 found in ancient graves. However, these are few, far and in-between. As far as I am aware, there are no moderns descending from these Neolithic samples that are unrelated to the majority of the branch.
There is no doubt that majority of branches arrived with Germanic speakers between late antiquity and early medieval. Slavic speakers are also not out of the question depending on the haplotype, the time period and its diversity. Germans and Slavs were neighbors for far longer periods than their trek to the Balkans. Some probably were Latinized too, as there were Gothic Foederati. The cluster in Albanians was probably previously Latin speaking. At least the most dominant cluster in the project.
I don't doubt that some I1 in the Balkans may be found that is unrelated to Germanic/Gothic peoples prior to late antiquity. I even mentioned 2 such branches found in Albanians that could represent the idea you have. The majority belong to the young branch I mentioned in Albanians and some Montenegrins.
Not sure you're math makes any sense. If you combine I2/R1a they are alot higher than I1 is. Assuming of course most of their haplotypes arrived with Slavs. So, no, I1 does not exceed R1a/I2a as a whole. Even in the North. I1 is actually a strong haplogroup for Serbs. Depending on region of course. Overall they get 8-10% I1 depending on the sources.
The breakdown is this. I only included Y3120 for I2a:
https://rrenjet.com/statistikat/
Rrenjet, which has the most samples:
Albanians as a whole have 15%(R1a/I2a) & 6.8%(I1)
Gegs have 11.9%(R1a/I2a) & 6.4%(I1)
Tosks have 20.3%(R1a/I2a) & 7.6%(I1)
Labs have 5.5%(R1a/I2a) & 13.2%(I1)
Cams have 35.3%(R1a/I2a) & 0%(I1)
Kosova has 13.8%(R1a/I2a) & 8.5%(I1)
Maqedoni has 15%(R1a/I2a) & 0%(I1)
Mali i Zi has 22.2%(R1a/I2a) & 0%(I1)
As you can see, only in the case of the Labs is I1 higher than R1a+I2a. Keep in mind, there are only 5 samples out of a small amount for Kosova in Rrenjet, and 14 total on Gjenetika that are I1. That could change as the region is sampled more. I wouldn't make any hard claims based off of 5-14 samples. A good many regions in Albania, Kosova, and Macedonia are still very under-sampled, or not sampled at all. In a small batch for Kosova recently done for Opoja on Rrenjet, there were no I1 and quite a bit of I-Y3120.
As for Gjenetika, which reports a number of samples that were migrated to Rrenjet, the statistics are as follows.(some of which may be duplicates)
http://www.gjenetika.com/statistikat/
Albanians as a whole have 12.6%(R1a/I2a) & 7.6%(I1)
Gegs have 8.6%(R1a/I2a) & 7.9%(I1)
Tosks have 20.1%(R1a/I2a) & 7.3%(I1)
Kosova has 7.5%(R1a/I2a) & 7.4%(I1)
Maqedoni has 10%(R1a/I2a) & 4%(I1)
Mali i Zi has 6.3%(R1a/I2a) & 3.1%(I1)
You can check the I1 tree yourself. None of them beside the 2 branches I mentioned are native to the Balkans. They mostly arrived with Germanic speakers who were Latinized, and minimally maybe some Slavicized Goths could have carried some branches into Serbs, as TMRCA's would suggest.
https://www.yfull.com/tree/I1/