Taranis, I fully agree. When Celts or IE in general arrived they possessed fully complete EI grammar. The best attestation is Latin.
I just don't believe that simplification of otherwise fully functional language would develop by itself, fashion wise, without strong stimuli.
Without good historic sources about spoken languages from the past it's hard to be 100% sure, which way it went.
In this case I was analyzing, known better, recent event from middle ages trying to extrapolate them on earlier events.
Take Slavs for example. During their expansion, in middle of first millennium AD, they were warriors-farmers. Whatever they conquered they surely settled in fertile areas, around local villages, fully mixing and exposing their language. The depopulation that happened in central Europe, for whatever reason, certainly helped at this time. Almost all Slavic nations now have full almost original Slavic grammar, with exception of Macedonia and Bulgaria.
If we consider the above to be true, then case of Goths is obviously on opposite side of spectrum. Wherever they went they didn't leave much of language influence, except some vocabulary, definitely not in grammar or pronunciation. They didn't leave much of genetic imprint either, though they ruled huge areas in Europe, and for hundreds of years. Even if they started as farmers around year 0, they must have changed quickly in just warrior class. Warrior class doesn't mix much with local villagers. They rather live on higher grounds, in forts or castles, big urban centers, overseeing population more than mingling and teaching their language to locals. They might have been the precursor of feudalism.
Looking at Iberian Celtic situation my guess is that something in between, two former examples, happened. Locals were mostly non IE speakers. Celts managed to mingle enough with locals, and for long enough, that locals took over Celtic vocabulary, but influenced grammar simplification and retained most of their pronunciation (judging by similarities to Basque, rolling r being a big feature.
So, where the heck Basque survived the onslaught of Celts? They either came later just before Romans, or where hiding away in Pyrenees till Romans managed to quiet Celts. These two scenarios can explain lack of Celtic vocabulary in Basque, and adoption of Roman vocabulary. By accepting some Roman vocabulary, but retaining sounds and full grammar, I would conclude that Roman presence in Basque area was more managerial than en mass settlements of Latin speaking citizens in Basque villages, a la Slavs.